{"meta": {"count": 828, "db_response_time_ms": 90, "page": 1, "per_page": 100, "groups_count": null, "x_query": {"oql": "get works where retracted is (false)\n  and year is (2025)\n  and title-abstract has ((stemmed \"artificial intelligence\" OR stemmed \"deep learning\" OR stemmed \"language model\" OR stemmed \"large language model\" OR stemmed \"machine learning\" OR stemmed \"neural network\" OR stemmed \"vision language model\") AND (stemmed \"autoformalization\" OR stemmed \"formal proof\" OR stemmed \"mathematical reasoning\" OR stemmed \"theorem proving\"))\n  and type is (article or preprint or review);\nthen sample (834) of those works with seed (42)", "oqo": {"get_rows": "works", "filter_rows": [{"column_id": "is_retracted", "value": false}, {"column_id": "publication_year", "value": 2025}, {"join": "or", "filters": [{"column_id": "title_and_abstract.search", "value": "\"artificial intelligence\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"deep learning\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"language model\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"large language model\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"machine learning\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"neural network\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"vision language model\"", "operator": "has"}]}, {"join": "or", "filters": [{"column_id": "title_and_abstract.search", "value": "\"autoformalization\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"formal proof\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"mathematical reasoning\"", "operator": "has"}, {"column_id": "title_and_abstract.search", "value": "\"theorem proving\"", "operator": "has"}]}, {"join": "or", "filters": [{"column_id": "type", "value": "article"}, {"column_id": "type", "value": "preprint"}, {"column_id": "type", "value": "review"}]}], "sample": 834, "seed": "42"}, "url": "/works?filter=is_retracted:false,publication_year:2025,title_and_abstract.search:%22artificial intelligence%22|%22deep learning%22|%22language model%22|%22large language model%22|%22machine learning%22|%22neural network%22|%22vision language model%22,title_and_abstract.search:%22autoformalization%22|%22formal proof%22|%22mathematical reasoning%22|%22theorem proving%22,type:article|preprint|review&sample=834&seed=42&per_page=100&page=1"}, "cost_usd": 0.001}, "results": [{"id": "https://openalex.org/W4414536410", "doi": "https://doi.org/10.48550/arxiv.2508.13037", "title": "Can Large Models Teach Student Models to Solve Mathematical Problems Like Human Beings? A Reasoning Distillation Method via Multi-LoRA Interaction", "publication_year": 2025, "publication_date": "2025-08-18", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5073294147", "display_name": "X. Li", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Xinhe", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5029321864", "display_name": "Jiajun Liu", "orcid": "https://orcid.org/0000-0002-7871-0518", "observed_orcids": ["https://orcid.org/0000-0002-7871-0518"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Jiajun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100396091", "display_name": "Peng Wang", "orcid": "https://orcid.org/0000-0002-8745-320X", "observed_orcids": ["https://orcid.org/0000-0002-8745-320X", "https://orcid.org/0000-0001-5876-4650"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Peng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Recent": [0], "studies": [1], "have": [2, 9], "demonstrated": [3], "that": [4, 175, 225], "Large": [5], "Language": [6, 31], "Models": [7, 32], "(LLMs)": [8], "strong": [10], "mathematical": [11, 104, 214], "reasoning": [12, 28, 59, 105, 114, 215], "abilities": [13], "but": [14], "rely": [15], "on": [16, 63, 99, 131, 231], "hundreds": [17], "of": [18, 20, 26, 43, 91, 115, 188, 217], "billions": [19], "parameters.": [21], "To": [22], "tackle": [23], "the": [24, 100, 111, 132, 154, 172, 183, 186, 200, 213, 232, 238, 250], "challenge": [25], "poor": [27], "in": [29, 185], "Small": [30], "(SLMs),": [33], "existing": [34], "methods": [35], "typically": [36], "leverage": [37], "LLMs": [38], "to": [39, 53, 121, 147, 177, 181, 204], "generate": [40], "massive": [41], "amounts": [42], "data": [44], "for": [45, 103, 144], "cramming": [46], "training.": [47], "In": [48], "psychology,": [49], "they": [50], "are": [51, 197], "akin": [52], "System": [54, 72, 149], "1": [55], "thinking,": [56, 74, 92, 151], "which": [57, 140], "resolves": [58], "problems": [60], "rapidly": [61], "based": [62, 98], "experience": [64], "and": [65, 80, 113, 158, 190, 199, 246], "intuition.": [66], "However,": [67], "human": [68], "learning": [69], "also": [70], "requires": [71], "2": [73, 150], "where": [75, 235], "knowledge": [76, 167, 176], "is": [77], "first": [78], "acquired": [79], "then": [81], "reinforced": [82], "through": [83, 219], "practice.": [84], "Inspired": [85], "by": [86, 241], "such": [87], "two": [88], "distinct": [89], "modes": [90], "we": [93, 109, 126, 152, 192], "propose": [94], "a": [95, 128], "novel": [96], "method": [97, 240], "multi-LoRA": [101], "Interaction": [102], "Distillation": [106], "(LoRID).": [107], "First,": [108], "input": [110], "question": [112], "each": [116], "sample": [117], "into": [118], "an": [119, 136], "LLM": [120], "create": [122], "knowledge-enhanced": [123], "datasets.": [124], "Subsequently,": [125], "train": [127, 153], "LoRA": [129], "block": [130], "student": [133], "model": [134], "as": [135], "Intuitive": [137], "Reasoner": [138, 160], "(IR),": [139], "directly": [141], "generates": [142], "Chain-of-Thoughts": [143], "problem-solving.": [145], "Then,": [146], "imitate": [148], "Knowledge": [155], "Generator": [156], "(KG)": [157], "Deep": [159], "(DR),": [161], "respectively.": [162, 254], "The": [163], "former": [164], "outputs": [165, 196], "only": [166], "after": [168], "receiving": [169], "problems,": [170], "while": [171], "latter": [173], "uses": [174], "perform": [178], "reasoning.": [179], "Finally,": [180], "address": [182], "randomness": [184], "generation": [187], "IR": [189], "DR,": [191], "evaluate": [193], "whether": [194], "their": [195], "consistent,": [198], "inference": [201], "process": [202], "needs": [203], "be": [205], "iterated": [206], "if": [207], "not.": [208], "This": [209], "step": [210], "can": [211], "enhance": [212], "ability": [216], "SLMs": [218], "mutual": [220], "feedback.": [221], "Experimental": [222], "results": [223], "show": [224], "LoRID": [226], "achieves": [227], "state-of-the-art": [228], "performance,": [229], "especially": [230], "GSM8K": [233], "dataset,": [234], "it": [236], "outperforms": [237], "second-best": [239], "2.3%,": [242], "16.1%,": [243], "2.4%,": [244], "12.3%,": [245], "1.8%": [247], "accuracy": [248], "across": [249], "five": [251], "base": [252], "models,": [253]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.600600004196167, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.600600004196167, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11902", "display_name": "Intelligent Tutoring Systems and Adaptive Learning", "score": 0.06790000200271606, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.05180000141263008, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2508.13037", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2508.13037", "pdf_url": "https://arxiv.org/pdf/2508.13037", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4416349084", "doi": "https://doi.org/10.48550/arxiv.2510.24021", "title": "SelecTKD: Selective Token-Weighted Knowledge Distillation for LLMs", "publication_year": 2025, "publication_date": "2025-10-28", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5003944732", "display_name": "Haiduo Huang", "orcid": "https://orcid.org/0009-0000-6726-9221", "observed_orcids": ["https://orcid.org/0009-0000-6726-9221"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Huang, Haiduo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5068434200", "display_name": "Ji Young Song", "orcid": "https://orcid.org/0000-0002-5749-3032", "observed_orcids": ["https://orcid.org/0000-0002-5749-3032"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Song, Jiangcheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100330176", "display_name": "Yadong Zhang", "orcid": "https://orcid.org/0000-0003-2861-7662", "observed_orcids": ["https://orcid.org/0000-0003-2861-7662"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Yadong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5111732383", "display_name": "Pengju Ren", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ren, Pengju", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Knowledge": [0], "distillation": [1, 50], "(KD)": [2], "is": [3, 35], "a": [4, 46, 80, 134], "standard": [5], "route": [6], "to": [7, 58, 61, 63], "compress": [8], "Large": [9], "Language": [10], "Models": [11], "(LLMs)": [12], "into": [13], "compact": [14], "students,": [15], "yet": [16], "most": [17], "pipelines": [18], "uniformly": [19], "apply": [20, 64], "token-wise": [21], "loss": [22], "regardless": [23], "of": [24], "teacher": [25, 78], "confidence.": [26], "This": [27, 104], "indiscriminate": [28], "supervision": [29], "amplifies": [30], "noisy,": [31], "high-entropy": [32], "signals": [33], "and": [34, 89, 110, 123, 133, 142], "especially": [36], "harmful": [37], "under": [38], "large": [39], "teacher-student": [40], "capacity": [41], "gaps.": [42], "We": [43], "introduce": [44], "SelecTKD,": [45], "plug-and-play": [47], "Selective": [48], "Token-Weighted": [49], "framework": [51], "that": [52, 73], "shifts": [53], "the": [54, 69, 77], "focus": [55], "from": [56], "\"how": [57], "measure": [59], "divergence\"": [60], "\"where": [62], "learning\".": [65], "At": [66], "each": [67], "step,": [68], "student": [70], "proposes": [71], "tokens": [72, 93, 99], "are": [74, 100], "verified": [75], "by": [76, 118], "through": [79], "robust": [81], "propose-and-verify": [82], "procedure": [83], "with": [84, 108], "two": [85], "variants:": [86], "greedy": [87], "Top-k": [88], "non-greedy": [90], "Spec-k.": [91], "Accepted": [92], "receive": [94], "full": [95], "loss,": [96], "while": [97], "rejected": [98], "masked": [101], "or": [102, 152], "down-weighted.": [103], "objective-agnostic": [105], "design": [106], "works": [107], "on-": [109], "off-policy": [111], "data,": [112], "induces": [113], "an": [114], "implicit": [115], "curriculum": [116], "quantified": [117], "Token": [119], "Acceptance": [120], "Rate": [121], "(TAR),": [122], "stabilizes": [124], "optimization.": [125], "Across": [126], "instruction": [127], "following,": [128], "mathematical": [129], "reasoning,": [130], "code": [131], "generation,": [132], "VLM": [135], "setting,": [136], "SelecTKD": [137], "consistently": [138], "improves": [139], "strong": [140], "baselines": [141], "achieves": [143], "state-of-the-art": [144], "results": [145], "for": [146], "small": [147], "models": [148], "without": [149], "architectural": [150], "changes": [151], "extra": [153], "reference": [154], "models.": [155]}, "primary_topic": {"id": "https://openalex.org/T10036", "display_name": "Advanced Neural Network Applications", "score": 0.4587000012397766, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10036", "display_name": "Advanced Neural Network Applications", "score": 0.4587000012397766, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.3163999915122986, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.14169999957084656, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.24021", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.24021", "pdf_url": "https://arxiv.org/pdf/2510.24021", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417537343", "doi": "https://doi.org/10.48550/arxiv.2506.15894", "title": "Language Models can perform Single-Utterance Self-Correction of Perturbed Reasoning", "publication_year": 2025, "publication_date": "2025-06-18", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Silver, Sam", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Silver, Sam", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Sun, Jimin", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Sun, Jimin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Zhang, Ivan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Ivan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Hooker, Sara", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Hooker, Sara", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": null, "display_name": "Kim, Eddie", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kim, Eddie", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "have": [4], "demonstrated": [5], "impressive": [6], "mathematical": [7], "reasoning": [8, 25], "capabilities,": [9], "yet": [10], "their": [11, 62], "performance": [12], "remains": [13], "brittle": [14], "to": [15, 28, 56, 88], "minor": [16], "variations": [17], "in": [18, 116, 137], "problem": [19], "description": [20], "and": [21, 81, 91], "prompting": [22], "strategy.": [23], "Furthermore,": [24], "is": [26], "vulnerable": [27], "sampling-induced": [29], "errors": [30], "which": [31], "autoregressive": [32], "models": [33, 80], "must": [34], "primarily": [35], "address": [36], "using": [37], "self-correction": [38, 45, 73, 111], "via": [39], "additionally-generated": [40], "tokens.": [41], "To": [42], "better": [43], "understand": [44], "capabilities": [46, 112], "of": [47, 64, 78, 93, 121, 132], "recent": [48, 126], "models,": [49], "we": [50], "conduct": [51], "experiments": [52], "measuring": [53], "models'": [54], "ability": [55, 123], "self-correct": [57], "synthetic": [58], "perturbations": [59], "introduced": [60], "into": [61], "Chain": [63], "Thought": [65], "(CoT)": [66], "reasoning.": [67], "We": [68], "observe": [69], "robust": [70], "single-utterance": [71], "intrinsic": [72, 110], "behavior": [74], "across": [75], "a": [76], "range": [77], "open-weight": [79], "datasets,": [82], "ranging": [83], "from": [84], "subtle,": [85], "implicit": [86], "corrections": [87, 92], "explicit": [89], "acknowledgments": [90], "errors.": [94], "Our": [95], "findings": [96], "suggest": [97], "that": [98, 125], "LLMs,": [99], "including": [100], "those": [101], "not": [102], "finetuned": [103], "for": [104], "long": [105], "CoT,": [106], "may": [107], "possess": [108], "stronger": [109], "than": [113], "commonly": [114], "shown": [115], "the": [117], "literature.": [118], "The": [119], "presence": [120], "this": [122], "suggests": [124], "\"reasoning\"": [127], "model": [128], "work": [129], "involves": [130], "amplification": [131], "traits": [133], "already": [134], "meaningfully": [135], "present": [136], "models.": [138]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7705000042915344, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7705000042915344, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.11559999734163284, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.019700000062584877, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2506.15894", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2506.15894", "pdf_url": "https://arxiv.org/pdf/2506.15894", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4416125299", "doi": "https://doi.org/10.48550/arxiv.2504.04150", "title": "Reasoning on Multiple Needles In A Haystack", "publication_year": 2025, "publication_date": "2025-04-05", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5120394271", "display_name": "Yidong Wang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Wang, Yidong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"The": [0], "Needle": [1], "In": [2, 42, 97], "A": [3, 43], "Haystack": [4, 44], "(NIAH)": [5], "task": [6], "has": [7], "been": [8], "widely": [9], "used": [10], "to": [11, 70], "evaluate": [12], "the": [13, 39, 72, 89, 102, 121, 124, 127, 131, 141, 163, 171, 177], "long-context": [14], "question-answering": [15], "capabilities": [16], "of": [17, 54, 74, 126, 179], "Large": [18], "Language": [19], "Models": [20], "(LLMs).": [21], "However,": [22], "its": [23, 29], "reliance": [24], "on": [25, 136, 190], "simple": [26], "retrieval": [27, 145], "limits": [28], "effectiveness.": [30], "To": [31], "address": [32, 71], "this": [33, 64, 98, 137, 180], "limitation,": [34], "recent": [35], "studies": [36], "have": [37], "introduced": [38], "Multiple": [40], "Needles": [41], "Reasoning": [45], "(MNIAH-R)": [46], "task,": [47], "which": [48, 168], "incorporates": [49], "supporting": [50], "documents": [51], "(Multiple": [52], "needles)": [53], "multi-hop": [55], "reasoning": [56, 147, 185], "tasks": [57], "into": [58, 144], "a": [59, 151, 160], "distracting": [60], "context": [61, 94], "(Haystack}).": [62], "Despite": [63], "advancement,": [65], "existing": [66], "approaches": [67], "still": [68], "fail": [69], "issue": [73], "models": [75], "providing": [76], "direct": [77], "answers": [78], "from": [79], "internal": [80], "knowledge,": [81], "and": [82, 111, 146, 149], "they": [83], "do": [84], "not": [85], "explain": [86], "or": [87], "mitigate": [88, 170], "decline": [90], "in": [91, 123, 183], "accuracy": [92], "as": [93, 130], "length": [95, 125, 133], "increases.": [96, 134], "paper,": [99], "we": [100, 112, 139, 175], "tackle": [101], "memory-based": [103], "answering": [104], "problem": [105], "by": [106, 120], "filtering": [107], "out": [108], "direct-answer": [109], "questions,": [110], "reveal": [113], "that": [114], "performance": [115, 172, 189], "degradation": [116], "is": [117], "primarily": [118], "driven": [119], "reduction": [122], "thinking": [128, 142, 166], "process": [129, 143], "input": [132], "Building": [135], "insight,": [138], "decompose": [140], "stages": [148], "introduce": [150], "reflection": [152], "mechanism": [153], "for": [154], "multi-round": [155], "extension.": [156], "We": [157], "also": [158], "train": [159], "model": [161], "using": [162], "generated": [164], "iterative": [165], "process,": [167], "helps": [169], "degradation.": [173], "Furthermore,": [174], "demonstrate": [176], "application": [178], "retrieval-reflection": [181], "capability": [182], "mathematical": [184], "scenarios,": [186], "improving": [187], "GPT-4o's": [188], "AIME2024.": [191]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.9301999807357788, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.9301999807357788, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.028699999675154686, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10286", "display_name": "Information Retrieval and Search Behavior", "score": 0.021299999207258224, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2504.04150", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2504.04150", "pdf_url": "https://arxiv.org/pdf/2504.04150", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417257970", "doi": "https://doi.org/10.48550/arxiv.2506.08935", "title": "Can A Gamer Train A Mathematical Reasoning Model?", "publication_year": 2025, "publication_date": "2025-06-10", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5103066975", "display_name": "Andrew Young Shin", "orcid": "https://orcid.org/0000-0001-6713-0609", "observed_orcids": ["https://orcid.org/0000-0001-6713-0609"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Shin, Andrew", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"While": [0], "large": [1], "language": [2], "models": [3, 93], "(LLMs)": [4], "have": [5, 24], "achieved": [6], "remarkable": [7], "performance": [8, 87], "in": [9, 97], "various": [10], "tasks": [11], "including": [12], "mathematical": [13, 55, 72, 89, 107], "reasoning,": [14], "their": [15], "development": [16], "typically": [17], "demands": [18], "prohibitive": [19], "computational": [20], "resources.": [21], "Recent": [22], "advancements": [23], "reduced": [25], "costs": [26], "for": [27], "training": [28], "capable": [29], "models,": [30], "yet": [31], "even": [32], "these": [33], "approaches": [34], "rely": [35], "on": [36, 75, 88], "high-end": [37], "hardware": [38], "clusters.": [39], "In": [40], "this": [41], "paper,": [42], "we": [43, 67], "demonstrate": [44], "that": [45, 82, 105], "a": [46, 53, 69], "single": [47], "average": [48], "gaming": [49], "GPU": [50], "can": [51], "train": [52, 68], "solid": [54], "reasoning": [56, 73, 90, 108], "model,": [57], "by": [58], "integrating": [59], "reinforcement": [60], "learning": [61], "and": [62], "memory": [63, 81], "optimization": [64], "techniques.": [65], "Specifically,": [66], "1.5B": [70], "parameter": [71], "model": [74], "RTX": [76], "3080": [77], "Ti": [78], "of": [79], "16GB": [80], "achieves": [83], "comparable": [84], "or": [85], "better": [86], "benchmarks": [91], "than": [92], "several": [94], "times": [95], "larger,": [96], "resource-constrained": [98], "environments.": [99], "Our": [100], "results": [101], "challenge": [102], "the": [103], "paradigm": [104], "state-of-the-art": [106], "necessitates": [109], "massive": [110], "infrastructure,": [111], "democratizing": [112], "access": [113], "to": [114], "high-performance": [115], "AI": [116], "research.": [117], "https://github.com/shinandrew/YouronMath.": [118]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.3303000032901764, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.3303000032901764, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.16279999911785126, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.13439999520778656, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2506.08935", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2506.08935", "pdf_url": "https://arxiv.org/pdf/2506.08935", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4407570748", "doi": "https://doi.org/10.48550/arxiv.2502.08680", "title": "Mathematical Reasoning in Large Language Models: Assessing Logical and Arithmetic Errors across Wide Numerical Ranges", "publication_year": 2025, "publication_date": "2025-02-12", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5116269531", "display_name": "Safal Shrestha", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I120250893", "display_name": "New York University Abu Dhabi", "ror": "https://ror.org/00e5k0821", "country_code": "AE", "type": "education", "lineage": ["https://openalex.org/I120250893", "https://openalex.org/I57206974"]}], "countries": ["AE"], "is_corresponding": false, "raw_author_name": "Shrestha, Safal", "raw_affiliation_strings": ["New York University Abu Dhabi"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "New York University Abu Dhabi", "institution_ids": ["https://openalex.org/I120250893"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5109709876", "display_name": "Minwu Kim", "orcid": "https://orcid.org/0009-0005-8104-2142", "observed_orcids": ["https://orcid.org/0009-0005-8104-2142"]}, "institutions": [{"id": "https://openalex.org/I120250893", "display_name": "New York University Abu Dhabi", "ror": "https://ror.org/00e5k0821", "country_code": "AE", "type": "education", "lineage": ["https://openalex.org/I120250893", "https://openalex.org/I57206974"]}], "countries": ["AE"], "is_corresponding": false, "raw_author_name": "Kim, Minwu", "raw_affiliation_strings": ["New York University Abu Dhabi"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "New York University Abu Dhabi", "institution_ids": ["https://openalex.org/I120250893"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5101801953", "display_name": "Keith W. Ross", "orcid": "https://orcid.org/0000-0003-2556-805X", "observed_orcids": ["https://orcid.org/0000-0003-2556-805X"]}, "institutions": [{"id": "https://openalex.org/I120250893", "display_name": "New York University Abu Dhabi", "ror": "https://ror.org/00e5k0821", "country_code": "AE", "type": "education", "lineage": ["https://openalex.org/I120250893", "https://openalex.org/I57206974"]}], "countries": ["AE"], "is_corresponding": false, "raw_author_name": "Ross, Keith", "raw_affiliation_strings": ["New York University Abu Dhabi"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "New York University Abu Dhabi", "institution_ids": ["https://openalex.org/I120250893"]}]}], "abstract_inverted_index": {"Mathematical": [0], "reasoning": [1, 39, 90, 120, 155], "in": [2, 59, 104, 119, 166], "Large": [3], "Language": [4], "Models": [5], "(LLMs)": [6], "is": [7], "often": [8], "evaluated": [9], "using": [10], "benchmarks": [11], "with": [12, 97, 121], "limited": [13], "numerical": [14, 57, 68, 112, 123, 164], "ranges,": [15], "failing": [16], "to": [17, 33, 62, 108], "reflect": [18], "real-world": [19], "problem-solving": [20], "across": [21, 66], "diverse": [22], "scales.": [23, 69], "Furthermore,": [24], "most": [25], "existing": [26], "evaluation": [27, 88, 151], "methods": [28], "only": [29], "compare": [30], "model": [31, 64], "outputs": [32], "ground-truth": [34], "answers,": [35], "obscuring": [36], "insights": [37], "into": [38], "processes.": [40], "To": [41], "address": [42], "these": [43], "limitations,": [44], "we": [45, 71], "introduce": [46], "GSM-Ranges,": [47], "a": [48, 73, 85, 101, 116, 149], "dataset": [49], "generator": [50], "derived": [51], "from": [52], "GSM8K": [53], "that": [54, 77], "systematically": [55], "perturbs": [56], "values": [58], "math": [60], "problems": [61], "assess": [63], "robustness": [65], "varying": [67], "Additionally,": [70], "propose": [72], "novel": [74], "grading": [75], "methodology": [76], "distinguishes": [78], "between": [79], "logical": [80, 105], "and": [81, 157], "non-logical": [82], "errors,": [83], "offering": [84], "more": [86], "precise": [87], "of": [89, 152], "processes": [91], "beyond": [92], "computational": [93], "accuracy.": [94], "Our": [95], "experiments": [96], "various": [98], "models": [99, 127], "reveal": [100], "significant": [102], "increase": [103], "error": [106], "rates-up": [107], "14": [109], "percentage": [110], "points-as": [111], "complexity": [113], "rises,": [114], "demonstrating": [115], "general": [117], "weakness": [118], "out-of-distribution": [122], "values.": [124], "Moreover,": [125], "while": [126], "demonstrate": [128], "high": [129], "accuracy": [130], "on": [131], "standalone": [132], "arithmetic": [133], "tasks,": [134], "their": [135], "performance": [136], "deteriorates": [137], "substantially": [138], "when": [139], "computations": [140], "are": [141], "embedded": [142], "within": [143], "word": [144], "problems.": [145], "These": [146], "findings": [147], "provide": [148], "comprehensive": [150], "LLMs'": [153], "mathematical": [154], "capabilities": [156], "inform": [158], "future": [159], "research": [160], "directions": [161], "for": [162], "improving": [163], "generalization": [165], "language": [167], "models.": [168]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7294999957084656, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7294999957084656, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.04859999939799309, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.03519999980926514, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2502.08680", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2502.08680", "pdf_url": "https://arxiv.org/pdf/2502.08680", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 1, "is_retracted": false}, {"id": "https://openalex.org/W4416365886", "doi": "https://doi.org/10.48550/arxiv.2508.00410", "title": "Co-rewarding: Stable Self-supervised RL for Eliciting Reasoning in Large Language Models", "publication_year": 2025, "publication_date": "2025-08-01", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5067204895", "display_name": "Zizhuo Zhang", "orcid": "https://orcid.org/0000-0003-1943-5963", "observed_orcids": ["https://orcid.org/0000-0003-1943-5963"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}, {"id": "https://openalex.org/I80804394", "display_name": "Southwest Baptist University", "ror": "https://ror.org/02dgbcj80", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I80804394"]}], "countries": ["HK", "US"], "is_corresponding": false, "raw_author_name": "Zhang, Zizhuo", "raw_affiliation_strings": ["TMLR Group , Department of Computer Science , Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TMLR Group , Department of Computer Science , Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987", "https://openalex.org/I80804394"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5106550627", "display_name": "Jianing Zhu", "orcid": "https://orcid.org/0000-0002-0100-5562", "observed_orcids": ["https://orcid.org/0000-0002-0100-5562"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}, {"id": "https://openalex.org/I80804394", "display_name": "Southwest Baptist University", "ror": "https://ror.org/02dgbcj80", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I80804394"]}], "countries": ["HK", "US"], "is_corresponding": false, "raw_author_name": "Zhu, Jianing", "raw_affiliation_strings": ["TMLR Group , Department of Computer Science , Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TMLR Group , Department of Computer Science , Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987", "https://openalex.org/I80804394"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Ge, Xinmu", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I183067930", "display_name": "Shanghai Jiao Tong University", "ror": "https://ror.org/0220qvk04", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I183067930"]}, {"id": "https://openalex.org/I4405279436", "display_name": "Shanghai Innovation Institute", "ror": "https://ror.org/01fatbw66", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I4405279436"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Ge, Xinmu", "raw_affiliation_strings": ["CMIC , Shanghai Jiao Tong University", "Shanghai Innovation Institute"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CMIC , Shanghai Jiao Tong University", "institution_ids": ["https://openalex.org/I183067930"]}, {"raw_affiliation_string": "Shanghai Innovation Institute", "institution_ids": ["https://openalex.org/I4405279436"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102705176", "display_name": "Zihua Zhao", "orcid": "https://orcid.org/0000-0001-8611-8656", "observed_orcids": ["https://orcid.org/0000-0001-8611-8656", "https://orcid.org/0009-0003-2915-6893"]}, "institutions": [{"id": "https://openalex.org/I183067930", "display_name": "Shanghai Jiao Tong University", "ror": "https://ror.org/0220qvk04", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I183067930"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Zhao, Zihua", "raw_affiliation_strings": ["CMIC , Shanghai Jiao Tong University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CMIC , Shanghai Jiao Tong University", "institution_ids": ["https://openalex.org/I183067930"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101267097", "display_name": "Zhanke Zhou", "orcid": "https://orcid.org/0009-0008-1663-9358", "observed_orcids": ["https://orcid.org/0009-0008-1663-9358"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}, {"id": "https://openalex.org/I80804394", "display_name": "Southwest Baptist University", "ror": "https://ror.org/02dgbcj80", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I80804394"]}], "countries": ["HK", "US"], "is_corresponding": false, "raw_author_name": "Zhou, Zhanke", "raw_affiliation_strings": ["TMLR Group , Department of Computer Science , Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TMLR Group , Department of Computer Science , Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987", "https://openalex.org/I80804394"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5108047728", "display_name": "Xuan Li", "orcid": "https://orcid.org/0000-0002-9172-4105", "observed_orcids": ["https://orcid.org/0000-0002-9172-4105"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}, {"id": "https://openalex.org/I80804394", "display_name": "Southwest Baptist University", "ror": "https://ror.org/02dgbcj80", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I80804394"]}], "countries": ["HK", "US"], "is_corresponding": false, "raw_author_name": "Li, Xuan", "raw_affiliation_strings": ["TMLR Group , Department of Computer Science , Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TMLR Group , Department of Computer Science , Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987", "https://openalex.org/I80804394"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5003764862", "display_name": "Feng Xiao", "orcid": "https://orcid.org/0000-0002-2895-1727", "observed_orcids": ["https://orcid.org/0000-0002-2895-1727"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}, {"id": "https://openalex.org/I80804394", "display_name": "Southwest Baptist University", "ror": "https://ror.org/02dgbcj80", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I80804394"]}], "countries": ["HK", "US"], "is_corresponding": false, "raw_author_name": "Feng, Xiao", "raw_affiliation_strings": ["TMLR Group , Department of Computer Science , Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TMLR Group , Department of Computer Science , Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987", "https://openalex.org/I80804394"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102922412", "display_name": "Jiangchao Yao", "orcid": "https://orcid.org/0000-0001-6115-5194", "observed_orcids": ["https://orcid.org/0000-0001-6115-5194"]}, "institutions": [{"id": "https://openalex.org/I183067930", "display_name": "Shanghai Jiao Tong University", "ror": "https://ror.org/0220qvk04", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I183067930"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Yao, Jiangchao", "raw_affiliation_strings": ["CMIC , Shanghai Jiao Tong University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CMIC , Shanghai Jiao Tong University", "institution_ids": ["https://openalex.org/I183067930"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5101418359", "display_name": "Bo Han", "orcid": "https://orcid.org/0000-0002-8633-7661", "observed_orcids": ["https://orcid.org/0000-0002-8633-7661"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}, {"id": "https://openalex.org/I80804394", "display_name": "Southwest Baptist University", "ror": "https://ror.org/02dgbcj80", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I80804394"]}], "countries": ["HK", "US"], "is_corresponding": false, "raw_author_name": "Han, Bo", "raw_affiliation_strings": ["TMLR Group , Department of Computer Science , Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TMLR Group , Department of Computer Science , Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987", "https://openalex.org/I80804394"]}]}], "abstract_inverted_index": {"While": [0], "reinforcement": [1], "learning": [2], "with": [3, 134, 206, 221], "verifiable": [4], "rewards": [5], "(RLVR)": [6], "is": [7, 106, 124, 229], "effective": [8], "to": [9, 25, 41, 137, 148, 166], "improve": [10], "the": [11, 26, 43, 52, 58, 64, 68, 73, 150, 169], "reasoning": [12, 44, 157, 192], "ability": [13], "of": [14, 46, 75, 146, 152, 217], "large": [15], "language": [16], "models": [17], "(LLMs),": [18], "its": [19], "reliance": [20], "on": [21, 155, 187, 189, 197, 219], "human-annotated": [22], "labels": [23, 136], "leads": [24], "scaling": [27], "up": [28], "dilemma,": [29], "especially": [30, 194], "for": [31], "complex": [32], "tasks.": [33], "Recent": [34], "self-rewarding": [35, 182], "methods": [36], "investigate": [37], "a": [38, 81, 107, 125, 130, 215], "label-free": [39], "alternative": [40], "unlock": [42], "capabilities": [45], "LLMs,": [47], "yet": [48], "they": [49], "frequently": [50], "encounter": [51], "non-negligible": [53], "training": [54, 88, 153, 175], "collapse": [55, 154], "issue,": [56], "as": [57, 214], "single-view": [59], "supervision": [60, 93], "signal": [61], "easily": [62], "forms": [63], "self-consistent": [65], "illusion,": [66], "yielding": [67], "reward": [69, 112], "hacking.": [70], "Inspired": [71], "by": [72, 90, 184, 195], "success": [74], "self-supervised": [76, 83], "learning,": [77], "we": [78, 98], "propose": [79], "\\textit{Co-rewarding},": [80], "novel": [82], "RL": [84], "framework": [85], "that": [86, 110, 128], "improves": [87], "stability": [89], "seeking": [91], "complementary": [92], "from": [94, 114], "another": [95], "views.": [96], "Specifically,": [97], "instantiate": [99], "Co-rewarding": [100, 172, 200], "in": [101, 210], "two": [102], "ways:": [103], "(1)": [104], "\\textit{Co-rewarding-I}": [105], "data-side": [108], "instantiation": [109, 127], "derives": [111], "signals": [113], "contrastive": [115], "agreement": [116], "across": [117, 176], "semantically": [118], "analogous": [119], "questions;": [120], "and": [121, 179], "(2)": [122], "\\textit{Co-rewarding-II}": [123], "model-side": [126], "maintains": [129], "slowly-updated": [131], "reference": [132], "teacher": [133], "pseudo": [135], "realize": [138], "self-distillation.": [139], "Intuitively,": [140], "such": [141, 213], "instantiations": [142], "introduce": [143], "different": [144], "levels": [145], "discrepancy": [147], "increase": [149], "difficulty": [151], "trivial": [156], "solutions.": [158], "We": [159], "also": [160], "explore": [161], "their": [162], "orthogonally": [163], "combined": [164], "version": [165], "further": [167], "boost": [168], "performance.": [170], "Empirically,": [171], "exhibits": [173], "stable": [174], "various": [177], "setups,": [178], "outperforms": [180], "other": [181], "baselines": [183], "$+3.31\\%$": [185], "improvements": [186], "average": [188], "multiple": [190], "mathematical": [191], "benchmarks,": [193], "$+7.49\\%$": [196], "Llama-3.2-3B-Instruct.": [198], "Notably,": [199], "reaches": [201], "or": [202], "even": [203], "surpasses": [204], "RLVR": [205], "ground-truth": [207], "(GT)": [208], "label": [209], "several": [211], "cases,": [212], "Pass@$1$": [216], "$94.01\\%$": [218], "GSM8K": [220], "Qwen3-8B-Base": [222], "remarkably": [223], "higher": [224], "than": [225], "GT.": [226], "Our": [227], "code": [228], "released": [230], "at": [231], "https://github.com/tmlr-group/Co-rewarding.": [232]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5985999703407288, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5985999703407288, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.08420000225305557, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.05700000002980232, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2508.00410", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2508.00410", "pdf_url": "https://arxiv.org/pdf/2508.00410", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415315114", "doi": "https://doi.org/10.48550/arxiv.2510.06430", "title": "MathRobust-LV: Evaluation of Large Language Models' Robustness to Linguistic Variations in Mathematical Reasoning", "publication_year": 2025, "publication_date": "2025-10-07", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5050870104", "display_name": "Neeraja Kirtane", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kirtane, Neeraja", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5068173132", "display_name": "Yuvraj Khanna", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Khanna, Yuvraj", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5021548301", "display_name": "Peter Relan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Relan, Peter", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Large": [0], "language": [1], "models": [2, 106, 147, 166, 170, 176], "excel": [3], "on": [4, 97, 145], "math": [5, 9, 42], "benchmarks,": [6], "but": [7], "their": [8], "reasoning": [10, 196], "robustness": [11, 129, 187], "to": [12, 84, 157, 188], "linguistic": [13, 128, 189], "variation": [14, 190], "is": [15, 138, 191], "underexplored.": [16], "While": [17], "recent": [18], "work": [19], "increasingly": [20], "treats": [21], "high-difficulty": [22], "competitions": [23], "like": [24, 177], "the": [25, 28, 102, 155, 158], "IMO": [26], "as": [27, 141], "gold": [29], "standard": [30], "for": [31, 131, 164], "evaluating": [32], "reasoning,": [33], "we": [34, 69, 95], "believe": [35], "in": [36, 44, 110, 124, 198], "comprehensive": [37], "benchmarking": [38, 137], "of": [39], "high": [40], "school-level": [41], "problems": [43, 62, 100], "real": [45], "educational": [46, 111], "settings.": [47], "We": [48], "introduce": [49], "MathRobust-LV,": [50], "a": [51, 192], "test": [52], "set": [53], "and": [54, 80, 114], "evaluation": [55], "methodology": [56], "that": [57, 87, 149, 186], "mirrors": [58], "how": [59], "instructors": [60, 120], "rephrase": [61, 121], "across": [63], "assessments": [64], "while": [65, 76, 168], "keeping": [66], "difficulty": [67, 103], "constant:": [68], "change": [70], "surface": [71], "details": [72], "(names,": [73], "contexts,": [74], "variables)": [75], "preserving": [77], "numerical": [78], "structure": [79], "answers.": [81], "In": [82, 117], "contrast": [83], "prior": [85], "efforts": [86], "alter": [88], "problem": [89], "content": [90], "or": [91], "emphasize": [92], "IMO-level": [93], "tasks,": [94], "focus": [96], "high-school-level": [98], "dataset": [99], "at": [101], "level": [104], "where": [105], "are": [107, 162], "currently": [108], "deployed": [109], "settings:": [112], "tutoring": [113], "assessment": [115], "systems.": [116], "these": [118], "applications,": [119], "identical": [122], "concepts": [123], "varied": [125], "ways,": [126], "making": [127], "essential": [130], "reliable": [132], "deployment.": [133], "Although": [134], "MATH": [135], "data": [136], "often": [139], "regarded": [140], "saturated,": [142], "our": [143], "experiment": [144], "34": [146], "reveals": [148], "accuracy": [150], "declines": [151], "when": [152], "moving": [153], "from": [154], "baseline": [156], "variants.": [159], "These": [160], "drops": [161], "severe": [163], "smaller": [165], "(9-11%)": [167], "stronger": [169], "also": [171], "show": [172], "measurable": [173], "degradation.": [174], "Frontier": [175], "GPT-5,": [178], "Gemini-2.5pro": [179], "remain": [180], "comparatively": [181], "stable.": [182], "Our": [183], "results": [184], "highlight": [185], "fundamental": [193], "challenge,": [194], "exposing": [195], "vulnerabilities": [197], "models.": [199]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.447299987077713, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.447299987077713, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11902", "display_name": "Intelligent Tutoring Systems and Adaptive Learning", "score": 0.11289999634027481, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.07639999687671661, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.06430", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.06430", "pdf_url": "https://arxiv.org/pdf/2510.06430", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417086877", "doi": "https://doi.org/10.48550/arxiv.2512.04673", "title": "Cross-Task Benchmarking and Evaluation of General-Purpose and Code-Specific Large Language Models", "publication_year": 2025, "publication_date": "2025-12-04", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Das, Gunjan", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I11880225", "display_name": "National Institute of Technology Karnataka", "ror": "https://ror.org/01hz4v948", "country_code": "IN", "type": "education", "lineage": ["https://openalex.org/I11880225"]}], "countries": ["IN"], "is_corresponding": false, "raw_author_name": "Das, Gunjan", "raw_affiliation_strings": ["National Institute of Technology Karnataka India"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "National Institute of Technology Karnataka India", "institution_ids": ["https://openalex.org/I11880225"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5073904315", "display_name": "Paheli Bhattacharya", "orcid": "https://orcid.org/0000-0002-8412-8161", "observed_orcids": ["https://orcid.org/0000-0002-8412-8161"]}, "institutions": [{"id": "https://openalex.org/I4210151956", "display_name": "Robert Bosch (India)", "ror": "https://ror.org/04my8ty22", "country_code": "IN", "type": "company", "lineage": ["https://openalex.org/I4210151956", "https://openalex.org/I889804353"]}], "countries": ["IN"], "is_corresponding": false, "raw_author_name": "Bhattacharya, Paheli", "raw_affiliation_strings": ["Bosch Research and Technology Centre India"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Bosch Research and Technology Centre India", "institution_ids": ["https://openalex.org/I4210151956"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100699246", "display_name": "Rishabh Gupta", "orcid": "https://orcid.org/0000-0001-8634-6819", "observed_orcids": ["https://orcid.org/0000-0001-8634-6819"]}, "institutions": [{"id": "https://openalex.org/I4210151956", "display_name": "Robert Bosch (India)", "ror": "https://ror.org/04my8ty22", "country_code": "IN", "type": "company", "lineage": ["https://openalex.org/I4210151956", "https://openalex.org/I889804353"]}], "countries": ["IN"], "is_corresponding": false, "raw_author_name": "Gupta, Rishabh", "raw_affiliation_strings": ["Bosch Research and Technology Centre India"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Bosch Research and Technology Centre India", "institution_ids": ["https://openalex.org/I4210151956"]}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "have": [4, 26], "revolutionized": [5], "both": [6], "general": [7], "natural": [8, 85], "language": [9, 86], "processing": [10], "and": [11, 20, 39, 56, 70, 87, 104, 124], "domain-specific": [12], "applications": [13], "such": [14], "as": [15], "code": [16, 40, 82, 97], "synthesis,": [17], "legal": [18], "reasoning,": [19, 38, 69], "finance.": [21], "However,": [22], "while": [23], "prior": [24], "studies": [25], "explored": [27], "individual": [28], "model": [29, 75], "capabilities,": [30], "a": [31, 50], "systematic": [32], "cross-domain": [33], "comparison": [34], "that": [35, 93, 107], "unifies": [36], "linguistic,": [37], "understanding": [41], "abilities": [42], "remains": [43], "underexplored.": [44], "In": [45], "this": [46], "work,": [47], "we": [48, 73], "present": [49], "comprehensive": [51], "evaluation": [52], "of": [53], "five": [54], "general-purpose": [55, 120], "three": [57], "code-specific": [58], "state-of-the-art": [59], "LLMs": [60], "across": [61], "six": [62], "diverse": [63], "benchmarks": [64], "encompassing": [65], "linguistic": [66], "competence,": [67], "mathematical": [68], "trustworthiness.": [71], "Additionally,": [72], "analyze": [74], "behavior": [76], "on": [77], "the": [78], "CoNaLa": [79], "dataset": [80], "for": [81, 96, 109], "explanation,": [83], "comparing": [84], "code-specialized": [88], "LLMs.": [89], "Our": [90], "findings": [91], "reveal": [92], "models": [94, 121], "optimized": [95], "(e.g.,": [98], "CodeLLaMA": [99], "variants)": [100], "exhibit": [101], "strong": [102], "reasoning": [103], "syntactic": [105], "precision,": [106], "even": [108], "non-coding": [110], "tasks": [111], "can": [112], "show": [113], "measurable": [114], "performance": [115], "gains,": [116], "in": [117], "contrast": [118], "to": [119], "like": [122], "Mistral-7B": [123], "Llama-3-8B.": [125]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7904999852180481, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7904999852180481, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10260", "display_name": "Software Engineering Research", "score": 0.15719999372959137, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.0364999994635582, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2512.04673", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2512.04673", "pdf_url": "https://arxiv.org/pdf/2512.04673", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4416445771", "doi": "https://doi.org/10.36227/techrxiv.176369747.70179230/v1", "title": "Efficient Long Chain-of-Thought Elicitation through Synthetic Data Generation and Targeted Fine-Tuning", "publication_year": 2025, "publication_date": "2025-11-21", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5042342823", "display_name": "Abdulloh Abdulloh", "orcid": "https://orcid.org/0009-0003-7300-5638", "observed_orcids": ["https://orcid.org/0009-0003-7300-5638"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Abdulloh Abdulloh", "raw_affiliation_strings": ["Independent Researcher"], "raw_orcid": "https://orcid.org/0009-0003-7300-5638", "affiliations": [{"raw_affiliation_string": "Independent Researcher", "institution_ids": []}]}], "abstract_inverted_index": {"Recent": [0], "advances": [1], "in": [2, 124, 270, 314], "large": [3, 107], "language": [4, 126], "model": [5, 249], "(LLM)": [6], "reasoning": [7, 18, 122, 155, 207, 213, 278], "have": [8], "demonstrated": [9], "that": [10, 74, 147, 189, 223, 292], "scaling": [11], "inference-time": [12], "compute": [13], "through": [14, 103], "long": [15, 31, 80, 87, 120, 149, 182, 284, 316], "chain-of-thought": [16], "(CoT)": [17], "can": [19], "significantly": [20], "improve": [21], "performance": [22, 230, 286], "on": [23, 36, 79, 211, 252, 257], "complex": [24], "tasks.": [25], "However,": [26], "current": [27], "approaches": [28], "for": [29, 118, 262, 283, 299], "training": [30, 89, 133], "CoT": [32, 81, 88, 121, 150, 164, 183, 285, 317], "capabilities": [33, 123], "rely": [34], "heavily": [35], "either": [37], "(1)": [38, 169], "expensive": [39], "reinforcement": [40], "learning": [41, 203, 295], "(RL)": [42], "from": [43, 53, 106, 158, 172], "base": [44], "models": [45, 56, 127, 306], "requiring": [46, 235], "massive": [47], "computational": [48, 238], "resources,": [49], "or": [50, 134], "(2)": [51, 185], "distillation": [52], "proprietary": [54, 244], "frontier": [55, 137], "like": [57], "OpenAI\u2019s": [58], "o1,": [59], "which": [60, 277], "is": [61, 297], "neither": [62], "reproducible": [63, 312], "nor": [64], "accessible": [65, 315], "to": [66, 94, 136, 178, 204, 231, 243, 260, 310], "most": [67, 281], "researchers.": [68], "Furthermore,": [69], "existing": [70, 159], "work": [71], "has": [72], "shown": [73], "while": [75, 234], "supervised": [76], "fine-tuning": [77], "(SFT)": [78], "data": [82, 90, 144, 161, 226, 254], "improves": [83], "performance,": [84], "obtaining": [85], "high-quality": [86], "remains": [91], "challenging": [92], "due": [93], "limited": [95], "availability": [96], "and": [97, 162, 175, 194, 197, 219, 240, 265, 287, 305], "the": [98], "expense": [99], "of": [100, 272], "generating": [101], "it": [102], "rejection": [104], "sampling": [105], "teacher": [108], "models.": [109, 138, 245], "In": [110], "this": [111], "work,": [112], "we": [113, 221], "investigate": [114], "computationally": [115], "efficient": [116], "methods": [117, 233], "eliciting": [119], "smaller": [125], "(\u22648B": [128], "parameters)": [129], "without": [130], "extensive": [131], "RL": [132], "access": [135, 242], "We": [139, 274], "propose": [140], "a": [141], "novel": [142], "synthetic": [143, 225, 253], "generation": [145], "framework": [146], "constructs": [148], "trajectories": [151], "by": [152], "programmatically": [153], "composing": [154], "patterns": [156, 279], "extracted": [157], "web": [160], "short": [163], "examples.": [165], "Our": [166, 246], "approach": [167, 227, 296], "combines": [168], "pattern": [170], "mining": [171], "discussion": [173], "forums": [174], "educational": [176], "content": [177], "identify": [179], "naturally": [180], "occurring": [181], "structures,": [184], "rule-based": [186], "trajectory": [187], "construction": [188], "systematically": [190], "incorporates": [191], "self-correction,": [192], "branching,": [193], "verification": [195], "steps,": [196], "(3)": [198], "targeted": [199], "SFT": [200], "with": [201], "curriculum": [202, 294], "progressively": [205], "build": [206], "capabilities.": [208], "Through": [209], "experiments": [210], "mathematical": [212], "benchmarks": [214], "including": [215], "MATH,": [216], "AIME": [217], "2024,": [218], "TheoremQA,": [220], "demonstrate": [222], "our": [224, 293], "achieves": [228, 255], "comparable": [229], "distillation-based": [232], "10\u00d7": [236], "less": [237], "resources": [239], "no": [241], "best": [247], "7B": [248], "trained": [250], "solely": [251], "52.3%": [256], "MATH-500": [258], "(compared": [259], "54.1%": [261], "QwQ-distilled": [263], "baselines)": [264], "shows": [266], "emergent": [267], "self-correction": [268], "behaviors": [269], "23%": [271], "responses.": [273], "further": [275], "analyze": [276], "are": [280], "critical": [282], "provide": [288], "ablation": [289], "studies": [290], "demonstrating": [291], "essential": [298], "stable": [300], "training.": [301, 318], "All": [302], "code,": [303], "data,": [304], "will": [307], "be": [308], "released": [309], "facilitate": [311], "research": [313]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8090999722480774, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8090999722480774, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.04520000144839287, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.03060000017285347, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.36227/techrxiv.176369747.70179230/v1", "is_oa": true, "landing_page_url": "https://doi.org/10.36227/techrxiv.176369747.70179230/v1", "pdf_url": "https://www.techrxiv.org/doi/pdf/10.36227/techrxiv.176369747.70179230/v1", "source": null, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "acceptedVersion", "is_accepted": true, "is_published": false, "raw_source_name": null, "raw_type": "posted-content"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414587187", "doi": "https://doi.org/10.48550/arxiv.2505.19706", "title": "Error Typing for Smarter Rewards: Improving Process Reward Models with Error-Aware Hierarchical Supervision", "publication_year": 2025, "publication_date": "2025-05-26", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5107672084", "display_name": "Tej Deep Pala", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I152815399", "display_name": "Singapore University of Technology and Design", "ror": "https://ror.org/05j6fvn87", "country_code": "SG", "type": "education", "lineage": ["https://openalex.org/I152815399"]}], "countries": ["SG"], "is_corresponding": false, "raw_author_name": "Pala, Tej Deep", "raw_affiliation_strings": ["Singapore University of Technology and Design"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Singapore University of Technology and Design", "institution_ids": ["https://openalex.org/I152815399"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5088591587", "display_name": "Pritha Sharma", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I152815399", "display_name": "Singapore University of Technology and Design", "ror": "https://ror.org/05j6fvn87", "country_code": "SG", "type": "education", "lineage": ["https://openalex.org/I152815399"]}], "countries": ["SG"], "is_corresponding": false, "raw_author_name": "Sharma, Panshul", "raw_affiliation_strings": ["Singapore University of Technology and Design"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Singapore University of Technology and Design", "institution_ids": ["https://openalex.org/I152815399"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5112033266", "display_name": "Amir Zadeh", "orcid": "https://orcid.org/0000-0002-5297-3571", "observed_orcids": ["https://orcid.org/0000-0002-5297-3571"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zadeh, Amir", "raw_affiliation_strings": ["Lambda Labs"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Lambda Labs", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100642796", "display_name": "Chuan Li", "orcid": "https://orcid.org/0000-0001-5101-4048", "observed_orcids": ["https://orcid.org/0000-0001-5101-4048"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Chuan", "raw_affiliation_strings": ["Lambda Labs"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Lambda Labs", "institution_ids": []}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5033376109", "display_name": "Soujanya Poria", "orcid": "https://orcid.org/0000-0001-6924-7931", "observed_orcids": ["https://orcid.org/0000-0001-6924-7931"]}, "institutions": [{"id": "https://openalex.org/I152815399", "display_name": "Singapore University of Technology and Design", "ror": "https://ror.org/05j6fvn87", "country_code": "SG", "type": "education", "lineage": ["https://openalex.org/I152815399"]}], "countries": ["SG"], "is_corresponding": false, "raw_author_name": "Poria, Soujanya", "raw_affiliation_strings": ["Singapore University of Technology and Design"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Singapore University of Technology and Design", "institution_ids": ["https://openalex.org/I152815399"]}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2, 22, 29], "(LLMs)": [3], "are": [4], "prone": [5], "to": [6, 35, 65, 114], "hallucination,": [7], "especially": [8], "during": [9], "multi-hop": [10], "and": [11, 54, 83, 139], "reasoning-intensive": [12], "tasks": [13], "such": [14], "as": [15], "mathematical": [16, 154], "problem": [17], "solving.": [18], "While": [19], "Outcome": [20], "Reward": [21, 28], "verify": [23], "only": [24, 143], "final": [25], "answers,": [26], "Process": [27], "(PRMs)": [30], "score": [31], "each": [32, 58], "intermediate": [33], "step": [34, 67], "steer": [36], "generation": [37], "toward": [38], "coherent": [39], "solutions.": [40], "We": [41], "introduce": [42], "PathFinder-PRM,": [43, 71], "a": [44, 74, 95, 124], "novel": [45], "hierarchical,": [46], "error-aware": [47], "discriminative": [48], "PRM": [49], "that": [50, 135], "first": [51], "classifies": [52], "math": [53], "consistency": [55], "errors": [56], "at": [57], "step,": [59], "then": [60], "combines": [61], "these": [62], "fine-grained": [63, 145], "signals": [64], "estimate": [66], "correctness.": [68], "To": [69], "train": [70], "we": [72], "construct": [73], "400K-sample": [75], "dataset": [76], "by": [77], "enriching": [78], "the": [79, 102, 129], "human-annotated": [80], "PRM800K": [81], "corpus": [82], "RLHFlow": [84], "Mistral": [85], "traces": [86], "with": [87, 156], "three-dimensional": [88], "step-level": [89], "labels.": [90], "On": [91], "PRMBench,": [92], "PathFinder-PRM": [93], "achieves": [94], "new": [96], "state-of-the-art": [97], "PRMScore": [98], "of": [99], "67.7,": [100], "outperforming": [101], "prior": [103], "best": [104], "(65.5)": [105], "while": [106], "using": [107], "3": [108], "times": [109], "less": [110], "data.": [111], "When": [112], "applied": [113], "reward": [115, 140], "guided": [116], "greedy": [117], "search,": [118], "our": [119], "model": [120], "yields": [121], "prm@8": [122], "48.3,": [123], "+1.5": [125], "point": [126], "gain": [127], "over": [128], "strongest": [130], "baseline.": [131], "These": [132], "results": [133], "demonstrate": [134], "decoupled": [136], "error": [137, 146], "detection": [138, 147], "estimation": [141], "not": [142], "boost": [144], "but": [148], "also": [149], "substantially": [150], "improve": [151], "end-to-end,": [152], "reward-guided": [153], "reasoning": [155], "greater": [157], "data": [158], "efficiency.": [159]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.6460000276565552, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.6460000276565552, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.17790000140666962, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.01510000042617321, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2505.19706", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2505.19706", "pdf_url": "https://arxiv.org/pdf/2505.19706", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by-sa", "license_id": "https://openalex.org/licenses/cc-by-sa", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417467424", "doi": "https://doi.org/10.48550/arxiv.2512.14252", "title": "G\u00f6del's Poetry", "publication_year": 2025, "publication_date": "2025-12-16", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Davis, Kelly J.", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Davis, Kelly J.", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Formal,": [0], "automated": [1], "theorem": [2, 22], "proving": [3], "has": [4], "long": [5], "been": [6], "viewed": [7], "as": [8, 125], "a": [9, 17, 50, 80], "challenge": [10], "to": [11, 20, 112, 141], "artificial": [12], "intelligence.": [13], "We": [14], "introduce": [15], "here": [16], "new": [18], "approach": [19], "computer": [21], "proving,": [23], "one": [24], "that": [25, 53], "employs": [26], "specialized": [27], "language": [28, 143], "models": [29, 46, 144], "for": [30], "Lean4": [31], "proof": [32, 58, 70, 116], "generation": [33], "combined": [34], "with": [35, 105, 147], "recursive": [36, 69, 115], "decomposition": [37, 60], "of": [38, 61, 73, 100], "difficult": [39, 62], "theorems": [40, 63], "into": [41, 64], "simpler": [42, 65], "entailing": [43, 66], "propositions.": [44, 75], "These": [45], "are": [47], "coordinated": [48], "through": [49], "multi-agent": [51], "architecture": [52], "orchestrates": [54], "autoformalization": [55], "(if": [56], "required),": [57], "generation,": [59], "propositions,": [67], "and": [68, 130, 145], "(and/or": [71], "decomposition)": [72], "these": [74], "Without": [76], "decomposition,": [77, 87], "we": [78], "achieve": [79], "90.4%": [81], "pass": [82], "rate": [83], "on": [84, 123], "miniF2F.": [85], "With": [86], "this": [88], "is": [89, 120], "significantly": [90], "improved.": [91], "A": [92], "key": [93], "technical": [94], "contribution": [95], "lies": [96], "in": [97], "our": [98], "extension": [99, 146], "the": [101, 131], "Kimina": [102], "Lean": [103], "Server": [104], "abstract": [106], "syntax": [107], "tree": [108], "(AST)": [109], "parsing": [110], "capabilities": [111], "facilitate": [113], "automated,": [114], "decomposition.": [117], "The": [118], "system": [119], "made": [121], "available": [122], "PyPI": [124], "goedels-poetry": [126], "(at": [127, 135], "https://pypi.org/project/goedels-poetry": [128], "),": [129], "open-source": [132], "implementation": [133], "KellyJDavis/goedels-poetry": [134], "https://github.com/KellyJDavis/goedels-poetry": [136], ")": [137], "facilitates": [138], "both": [139], "adaptation": [140], "alternative": [142], "custom": [148], "functionality.": [149]}, "primary_topic": {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.7821999788284302, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.7821999788284302, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.12129999697208405, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.01640000008046627, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2512.14252", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2512.14252", "pdf_url": "https://arxiv.org/pdf/2512.14252", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414836590", "doi": "https://doi.org/10.48550/arxiv.2502.15969", "title": "Forgotten Polygons: Multimodal Large Language Models are Shape-Blind", "publication_year": 2025, "publication_date": "2025-02-21", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Rudman, William", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Rudman, William", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5019199746", "display_name": "Michal Golovanevsky", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Golovanevsky, Michal", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5040676689", "display_name": "Amir Bar", "orcid": "https://orcid.org/0000-0001-8066-0495", "observed_orcids": ["https://orcid.org/0000-0001-8066-0495"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Bar, Amir", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5091920997", "display_name": "Vedant Palit", "orcid": "https://orcid.org/0009-0008-5202-8140", "observed_orcids": ["https://orcid.org/0009-0008-5202-8140"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Palit, Vedant", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5001226970", "display_name": "Yann LeCun", "orcid": "https://orcid.org/0000-0002-1992-2684", "observed_orcids": ["https://orcid.org/0000-0002-1992-2684"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "LeCun, Yann", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5014921416", "display_name": "Carsten Eickhoff", "orcid": "https://orcid.org/0000-0001-9895-4061", "observed_orcids": ["https://orcid.org/0000-0001-9895-4061"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Eickhoff, Carsten", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5070578596", "display_name": "Ritambhara Singh", "orcid": "https://orcid.org/0000-0002-7523-160X", "observed_orcids": ["https://orcid.org/0000-0002-7523-160X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Singh, Ritambhara", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Despite": [0], "strong": [1], "performance": [2, 25], "on": [3, 26, 93, 155], "vision-language": [4], "tasks,": [5], "Multimodal": [6], "Large": [7], "Language": [8], "Models": [9], "(MLLMs)": [10], "struggle": [11], "with": [12, 15, 67], "mathematical": [13, 143], "problem-solving,": [14], "both": [16, 113], "open-source": [17], "and": [18, 48, 88, 115, 178], "state-of-the-art": [19], "models": [20, 69], "falling": [21], "short": [22], "of": [23, 41, 85, 112, 125], "human": [24], "visual-math": [27], "benchmarks.": [28], "To": [29], "systematically": [30], "examine": [31], "visual-mathematical": [32], "reasoning": [33, 57, 144, 171], "in": [34, 64, 74, 150, 172], "MLLMs,": [35], "we": [36, 133], "(1)": [37], "evaluate": [38], "their": [39], "understanding": [40], "geometric": [42], "primitives,": [43], "(2)": [44], "test": [45], "multi-step": [46, 142], "reasoning,": [47], "(3)": [49], "explore": [50], "a": [51], "potential": [52], "solution": [53], "to": [54, 108, 163], "improve": [55], "visual": [56, 130, 148, 186], "capabilities.": [58], "Our": [59, 165], "findings": [60, 166], "reveal": [61], "fundamental": [62], "shortcomings": [63], "shape": [65], "recognition,": [66], "top": [68], "achieving": [70], "under": [71], "50%": [72], "accuracy": [73, 154], "identifying": [75], "regular": [76], "polygons.": [77], "We": [78], "analyze": [79], "these": [80], "failures": [81], "through": [82], "the": [83, 110, 123], "lens": [84], "dual-process": [86], "theory": [87], "show": [89], "that": [90, 168], "MLLMs": [91, 106, 173], "rely": [92], "System": [94, 101, 169], "1": [95], "(intuitive,": [96], "memorized": [97], "associations)": [98], "rather": [99], "than": [100], "2": [102, 170], "(deliberate": [103], "reasoning).": [104], "Consequently,": [105], "fail": [107], "count": [109], "sides": [111, 126], "familiar": [114], "novel": [116], "shapes,": [117], "suggesting": [118], "they": [119], "have": [120], "neither": [121], "learned": [122], "concept": [124], "nor": [127], "effectively": [128], "process": [129], "inputs.": [131], "Finally,": [132], "propose": [134], "Visually": [135], "Cued": [136], "Chain-of-Thought": [137], "(VC-CoT)": [138], "prompting,": [139], "which": [140], "enhances": [141], "by": [145], "explicitly": [146], "referencing": [147], "annotations": [149], "diagrams,": [151], "boosting": [152], "GPT-4o's": [153], "an": [156, 175], "irregular": [157], "polygon": [158], "side-counting": [159], "task": [160], "from": [161], "7%": [162], "93%.": [164], "suggest": [167], "remains": [174], "open": [176], "problem,": [177], "visually-guided": [179], "prompting": [180], "is": [181], "essential": [182], "for": [183], "successfully": [184], "engaging": [185], "reasoning.": [187], "Code": [188], "available": [189], "at:": [190], "https://github.com/rsinghlab/Shape-Blind.": [191]}, "primary_topic": {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.9980000257492065, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.9980000257492065, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.0003000000142492354, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.00019999999494757503, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2502.15969", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2502.15969", "pdf_url": "https://arxiv.org/pdf/2502.15969", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 2, "is_retracted": false}, {"id": "https://openalex.org/W4416681990", "doi": "https://doi.org/10.48550/arxiv.2509.00125", "title": "Know When to Explore: Difficulty-Aware Certainty as a Guide for LLM Reinforcement Learning", "publication_year": 2025, "publication_date": "2025-08-29", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Li, Ang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Ang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5088634717", "display_name": "Zhihang Yuan", "orcid": "https://orcid.org/0000-0003-4259-4719", "observed_orcids": ["https://orcid.org/0000-0003-4259-4719"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yuan, Zhihang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101456454", "display_name": "Yang Zhang", "orcid": "https://orcid.org/0000-0003-3056-3069", "observed_orcids": ["https://orcid.org/0000-0003-3056-3069"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Yang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5037807288", "display_name": "Shouda Liu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Shouda", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5101431030", "display_name": "Yisen Wang", "orcid": "https://orcid.org/0000-0002-0181-2137", "observed_orcids": ["https://orcid.org/0000-0002-0181-2137"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Yisen", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Reinforcement": [0], "Learning": [1], "with": [2, 87], "Verifiable": [3], "Feedback": [4], "(RLVF)": [5], "has": [6], "become": [7], "a": [8, 33, 101], "key": [9], "technique": [10], "for": [11, 137, 152], "enhancing": [12], "the": [13, 46, 56, 112, 123, 141], "reasoning": [14, 47, 167], "abilities": [15], "of": [16, 74], "Large": [17], "Language": [18], "Models": [19], "(LLMs).": [20], "However,": [21], "its": [22], "reliance": [23], "on": [24, 45, 122, 164], "sparse,": [25], "outcome": [26], "based": [27, 121], "rewards,": [28], "which": [29], "only": [30, 182], "indicate": [31], "if": [32], "final": [34], "answer": [35], "is": [36, 143], "correct": [37], "or": [38], "not,": [39], "fails": [40], "to": [41, 109, 132], "provide": [42], "granular": [43], "guidance": [44], "process": [48], "itself.": [49], "This": [50], "limitation": [51], "hinders": [52], "efficient": [53], "learning,": [54], "as": [55], "model": [57, 142], "cannot": [58], "distinguish": [59], "between": [60], "high": [61, 150, 161], "quality": [62], "and": [63, 90], "inefficient": [64], "solutions,": [65], "nor": [66], "can": [67], "it": [68, 155], "learn": [69], "effectively": [70], "from": [71], "different": [72], "types": [73], "failures.": [75], "To": [76], "address": [77], "this,": [78], "we": [79], "observe": [80], "that": [81, 105, 172, 197], "an": [82, 134], "LLMs": [83], "self-certainty": [84], "often": [85], "correlates": [86], "task": [88, 118], "difficulty": [89, 119], "solution": [91], "quality.": [92], "We": [93], "introduce": [94], "Difficulty": [95], "Aware": [96], "Certainty": [97], "guided": [98], "Exploration": [99], "(DACE),": [100], "novel": [102], "RL": [103], "algorithm": [104], "leverages": [106], "this": [107, 130], "insight": [108], "dynamically": [110], "balance": [111], "exploration": [113, 147, 203], "exploitation": [114], "trade-off.": [115], "DACE": [116, 145, 173], "assesses": [117], "online": [120], "policys": [124], "success": [125], "rate.": [126], "It": [127], "then": [128], "uses": [129], "signal": [131], "modulate": [133], "intrinsic": [135], "reward:": [136], "difficult": [138], "tasks": [139], "where": [140], "struggling,": [144], "encourages": [146, 156], "by": [148, 159], "penalizing": [149], "certainty;": [151], "easier": [153], "tasks,": [154], "learning": [157], "efficiency": [158], "rewarding": [160], "certainty.": [162], "Experiments": [163], "challenging": [165], "mathematical": [166], "benchmarks": [168], "(AIME,": [169], "MATH)": [170], "show": [171], "significantly": [174], "outperforms": [175], "strong": [176], "baselines.": [177], "The": [178], "DACE-trained": [179], "models": [180], "not": [181], "achieve": [183], "higher": [184], "accuracy": [185], "but": [186], "also": [187], "demonstrate": [188], "more": [189], "robust": [190], "performance": [191], "when": [192], "scaling": [193], "test-time": [194], "compute,": [195], "validating": [196], "our": [198], "adaptive": [199], "approach": [200], "fosters": [201], "effective": [202], "without": [204], "sacrificing": [205], "precision.": [206]}, "primary_topic": {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.8105000257492065, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.8105000257492065, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.041600000113248825, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12072", "display_name": "Machine Learning and Algorithms", "score": 0.024700000882148743, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2509.00125", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2509.00125", "pdf_url": "https://arxiv.org/pdf/2509.00125", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7117363068", "doi": "https://doi.org/10.5281/zenodo.18057869", "title": "Naiver strokes and yang mills solution are the same and Hodge", "publication_year": 2025, "publication_date": "2025-12-25", "type": "article", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5120885336", "display_name": "Vanni Borz", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Borz, Vanni", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Super-Hodge": [0, 1261], "framework": [1, 135, 546, 1002, 1019, 1082, 1126], "is": [2, 21, 35, 182, 214, 230, 271, 277, 284, 289, 299, 400, 422, 447, 459, 480, 513, 613, 632, 674, 689, 732, 802, 816, 849, 913, 943, 957, 990, 1027, 1127, 1209, 1306], "a": [3, 23, 28, 36, 43, 57, 90, 95, 97, 104, 108, 113, 125, 132, 216, 227, 258, 301, 349, 371, 383, 387, 391, 423, 448, 472, 505, 615, 705, 725, 892, 909, 958, 962, 1031, 1136, 1172, 1212, 1243, 1252], "global": [4, 231, 397, 449], "diagnostic": [5, 38, 178, 398, 450, 894, 1060, 1137], "system,": [6], "in": [7, 17, 70, 111, 200, 232, 374, 402, 471, 489, 584, 634, 647, 851, 1046], "the": [8, 79, 134, 138, 142, 159, 162, 165, 169, 176, 224, 233, 250, 255, 269, 297, 354, 363, 375, 428, 457, 490, 552, 619, 653, 656, 666, 692, 718, 783, 794, 797, 924, 936, 981, 992, 996, 1034, 1047, 1075, 1089, 1140, 1199, 1260, 1280, 1318, 1329], "strongest": [9], "and": [10, 73, 173, 281, 319, 369, 470, 496, 527, 540, 560, 564, 583, 600, 624, 751, 756, 768, 856, 877, 998, 1056, 1193, 1327], "most": [11, 429], "ambitious": [12], "sense": [13, 234], "one": [14, 829], "can": [15, 53, 867, 1013], "give": [16], "applied": [18, 869], "mathematics.": [19], "It": [20, 77, 181, 189, 229, 239, 262, 288, 452, 501, 720, 1026], "not": [22, 716, 771, 944, 961, 1008, 1115], "complete": [24, 1102], "physical": [25, 50, 695], "theory,": [26, 917, 1132], "nor": [27], "TOE": [29], "(Theory": [30], "of": [31, 184, 235, 254, 268, 366, 396, 431, 478, 509, 531, 618, 629, 655, 668, 785, 796, 899, 906, 926, 935, 970, 978, 995, 1033, 1171, 1178, 1279], "Everything),": [32], "but": [33, 463, 762, 808, 1006], "it": [34, 248, 446, 464, 479, 965], "universal": [37, 430, 1022], "system": [39, 270, 298], "that": [40, 52, 219, 437, 1070, 1206], "works": [41, 190], "like": [42, 222], "\u201cglobal": [44], "check-up\u201d": [45], "for": [46, 353], "any": [47, 359, 476, 1270], "mathematical": [48, 355, 468, 507, 709], "or": [49, 65, 186, 193, 197, 206, 245, 390, 485, 773, 882, 1040, 1095, 1117], "object": [51], "be": [54, 868, 1015, 1053], "described": [55], "through": [56], "Hodge-type": [58], "Laplacian": [59, 144, 256, 620, 1218, 1224, 1231, 1281], "(i.e.,": [60], "almost": [61], "all": [62, 440, 1248], "geometric,": [63], "differential,": [64], "dissipative": [66, 601, 685], "systems": [67, 85], "we": [68, 346], "encounter": [69], "modern": [71], "mathematics": [72, 537, 699], "physics).": [74], "Why": [75], "\u201cglobal\u201d?": [76], "applies": [78], "same": [80, 139, 177, 364, 376, 491, 864], "protocol": [81], "to": [82, 380, 439, 723, 782, 792, 1029, 1149], "completely": [83], "different": [84], "Whether": [86], "you": [87, 220, 264, 357, 379, 455], "feed": [88], "it:": [89], "Navier-Stokes": [91], "solution": [92], "(vorticity": [93, 872], "as": [94, 129, 344, 652, 758, 805, 812, 891, 908, 1135], "2-form),": [96], "cohomology": [98], "class": [99], "on": [100, 131, 191, 226, 427, 548, 575, 1219, 1251], "an": [101, 728, 806, 967], "algebraic": [102, 307], "variety,": [103], "Yang-Mills": [105], "gauge": [106], "field,": [107], "metric": [109], "perturbation": [110], "gravity,": [112], "topological": [114, 1176], "data": [115], "signal": [116], "(point": [117], "cloud": [118], "\u2192": [119, 121, 296, 317, 322, 332, 335, 1216, 1222, 1227, 1237], "Vietoris-Rips": [120], "combinatorial": [122], "Laplacian),": [123, 436], "even": [124], "neural": [126], "network": [127], "(interpretable": [128], "flow": [130, 645, 1226], "graph),": [133], "always": [136, 373], "does": [137, 715, 1007], "thing:": [140], "builds": [141], "appropriate": [143], "\u0394": [145], "=": [146, 294, 555, 607, 670, 1296, 1322, 1335], "\u2207*\u2207": [147], "+": [148, 315, 557], "R,": [149], "projects": [150], "onto": [151, 845], "three": [152], "spectral": [153, 367, 541, 562, 567, 616, 759, 788, 841, 865, 911, 1090, 1239, 1245, 1254], "sectors": [154, 602], "(holding,": [155], "quasi-holding,": [156], "dissipative),": [157], "computes": [158], "fractions": [160], "\u0398,": [161], "near-zero": [163, 787, 838], "gap,": [164, 416], "kernel": [166, 654, 667, 1278], "count,": [167], "checks": [168], "5": [170], "emergence": [171], "rules,": [172], "finally": [174], "outputs": [175], "signature": [179], "\u211c(T).": [180], "independent": [183], "scale": [185], "specific": [187], "context": [188], "compact": [192, 576], "non-compact": [194], "manifolds,": [195, 577], "with": [196, 203, 257, 386, 467, 516, 535, 579, 665], "without": [198], "boundary,": [199], "arbitrary": [201], "dimension,": [202], "Euclidean,": [204], "Lorentzian,": [205], "discrete": [207, 252, 585, 832], "metrics.": [208], "The": [209, 545, 604, 711, 863, 887, 904, 951, 1000, 1063, 1081, 1124, 1157], "parameter": [210], "\u03b5": [211], ">": [212], "0": [213, 1155], "just": [215, 241], "\u201cdiagnostic": [217], "resolution\u201d": [218], "choose,": [221], "changing": [223], "frequency": [225], "radar.": [228], "\u201cthe": [236], "entire": [237, 251], "spectrum\u201d": [238], "doesn\u2019t": [240, 453], "look": [242], "at": [243, 442, 681, 1088], "low": [244], "high": [246], "modes:": [247], "covers": [249], "spectrum": [253, 621], "rigorous": [259], "orthogonal": [260], "tripartition.": [261], "tells": [263, 465], "simultaneously": [265], "how": [266, 275, 282, 475], "much": [267, 276, 283], "\u201cfrozen": [272], "forever\u201d": [273], "(T_H),": [274], "\u201calmost": [278], "frozen\u201d": [279], "(T_Q),": [280], "\u201cdraining": [285], "away\u201d": [286], "(T_D).": [287], "diagnostically": [290], "predictive": [291], "If": [292, 313, 330], "Emerges(T)": [293], "1": [295, 318, 333, 1196], "producing": [300], "stable,": [302], "coherent,": [303], "recognizable": [304], "structure": [305], "(vortex,": [306], "cycle,": [308], "massive-like": [309], "particle,": [310], "horizon,": [311], "etc.).": [312, 418], "\u0398_H": [314], "\u0398_Q": [316], "large": [320], "gap": [321, 879, 1174], "saturation,": [323], "inertial": [324], "holding,": [325, 483, 598], "possible": [326, 1304], "delayed": [327], "causal": [328, 663, 1330], "release.": [329], "\u0398_D": [331], "rapidly": [334], "thermal": [336], "chaos,": [337], "no": [338, 684], "persistent": [339], "structure.": [340], "In": [341, 764], "practice,": [342], "it\u2019s": [343], "if": [345], "had": [347], "invented": [348], "standardized": [350], "medical": [351], "exam": [352], "universe:": [356], "bring": [358], "\u201cpatient\u201d": [360], "(system),": [361], "run": [362], "set": [365], "tests,": [368], "get": [370], "report": [372], "format,": [377], "allowing": [378], "directly": [381], "compare": [382], "turbulent": [384], "fluid": [385], "Calabi-Yau": [388], "manifold": [389], "black": [392], "hole.": [393], "This": [394, 502, 627, 731, 800, 848, 1051, 1241, 1338], "level": [395], "unification": [399, 1079], "rare": [401], "mathematics:": [403], "usually": [404], "each": [405, 1207], "field": [406], "has": [407, 1188], "its": [408, 1189], "own": [409, 1190], "indicators": [410], "(Reynolds": [411], "number,": [412], "Betti": [413], "numbers,": [414], "mass": [415, 878, 1173], "complexity,": [417], "Here,": [419], "instead,": [420], "there": [421], "single": [424], "language,": [425, 1191], "founded": [426], "operators": [432, 1240], "(the": [433], "generalized": [434], "Hodge": [435, 553, 833, 1217, 1232, 1266], "speaks": [438], "fields": [441, 844], "once.": [443], "So": [444], "yes:": [445], "system.": [451], "tell": [454], "why": [456], "universe": [458], "made": [460], "this": [461, 765, 1004, 1018], "way,": [462, 474], "you,": [466], "precision": [469], "comparable": [473, 1253], "part": [477], "structuring": [481], "itself,": [482], "dissipating,": [484], "emerging": [486], "\u2014": [487, 1156, 1197, 1258, 1277, 1285, 1291, 1311], "everywhere,": [488], "way.": [492], "Mathematical": [493, 1100], "Verification,": [494], "Compatibility,": [495], "What": [497, 815, 942, 1012], "Is": [498], "Feasible": [499, 862], "Beyond": [500], "section": [503], "provides": [504], "genuine": [506], "verification": [508], "theSuper\u2013Hodge": [510], "framework:": [511], "what": [512, 523, 528], "fully": [514], "compatible": [515, 1130], "existing": [517, 536, 826], "mathematics,": [518], "whatis": [519], "already": [520], "feasible": [521, 817, 824, 903], "today,": [522], "requires": [524, 932, 966], "additional": [525], "structure,": [526], "iscurrently": [529], "out": [530], "reach.": [532], "1.": [533], "Compatibility": [534], "(A)": [538, 947], "Hodge\u2013Laplacian": [539], "projectors": [542, 568], "Fully": [543], "standard": [544, 633, 675, 698, 750], "relies": [547], "classical,": [549], "well-established": [550], "objects:": [551], "Laplacian\u0394=dd\\*+d\\*d\\Delta": [554], "dd^\\*": [556], "d^\\*d\u0394=dd\\*+d\\*d,": [558], "self-adjoint": [559], "nonnegative;": [561], "decomposition": [563, 605, 1262], "functional": [565], "calculus;": [566], "P\u2264\u03b5P_{\\le\\varepsilon}P\u2264\u03b5,": [569], "Pker\u2061\u0394P_{\\ker\\Delta}Pker\u0394.": [570], "These": [571], "are": [572, 592, 700, 749, 770, 776], "rigorously": [573, 1083], "defined": [574], "manifolds": [578], "boundary(with": [580], "suitable": [581], "conditions),": [582], "settings": [586], "(DEC,": [587, 835], "FEM).": [588], "No": [589, 694], "new": [590], "assumptions": [591, 696], "introduced": [593], "here.": [594], "(B)": [595, 973], "Decomposition": [596], "into": [597, 622, 929], "quasi-harmonic,": [599], "Legitimate": [603], "T=TH\u2295TQ\u2295TDT": [606, 1295], "T_H": [608, 1297], "\\oplus": [609, 611, 1298, 1300], "T_Q": [610, 1299], "T_DT=TH\u2295TQ\u2295TD": [612, 1301], "simply": [614], "partition": [617], "kernel,near-kernel,": [623], "high-frequency": [625], "modes.": [626], "type": [628], "low/high-frequency": [630], "splitting": [631, 1269], "PDE": [635, 660, 677], "analysis,": [636], "including:": [637], "Navier\u2013Stokes": [638, 871, 997, 1215], "(Littlewood\u2013Paley": [639], "theory),": [640], "Yang\u2013Mills": [641, 874, 1221], "heat": [642], "flow,": [643], "Ricci": [644, 881, 1225], "(implicitly": [646], "Perelman\u2019s": [648], "work).": [649], "(C)": [650], "Light": [651], "wave": [657, 1319], "operator": [658, 938, 1201], "Classical": [659], "theory": [661, 1267], "Identifying": [662], "propagation": [664], "\u25a1=\u2202t2\u2212c2\u0394\\Box": [669], "\\partial_t^2": [671, 1323], "-": [672, 1324], "c^2\\Delta\u25a1=\u2202t2\u2212c2\u0394": [673], "hyperbolic": [676], "theory:": [678], "solutions": [679], "propagate": [680], "finite": [682], "speed,": [683], "decay": [686, 1294], "occurs,": [687], "causality": [688, 1313], "enforced": [690], "by": [691, 1120, 1204, 1211, 1268], "operator.": [693], "beyond": [697, 1017], "required.": [701], "(D)": [702], "Visibility": [703], "via": [704], "reading": [706], "map": [707, 713], "Existing": [708], "practice": [710], "visibility": [712], "\u03a0\\Pi\u03a0": [714], "modify": [717], "dynamics.": [719], "serves": [721], "only": [722], "detectwhether": [724], "component": [726], "produces": [727], "observable": [729], "signal.": [730], "mathematically": [733, 803, 1128], "equivalent": [734], "to:": [735, 870], "trace": [736], "operators,": [737, 742], "scattering": [738, 916], "maps,": [739], "boundary": [740], "measurement": [741], "tools": [743], "from": [744], "inverse": [745], "problems.": [746], "Such": [747], "constructions": [748], "well": [752], "understood.": [753], "(E)": [754], "RH": [755, 767, 780, 1234], "BSD": [757, 769, 790, 1236], "readers": [760], "Interpretative": [761], "consistent": [763, 804, 914], "framework,": [766], "reformulated": [772], "proved.": [774], "They": [775], "used": [777], "asstructural": [778], "analogies:": [779], "corresponds": [781, 791], "absence": [784], "hidden": [786], "bands,": [789], "counting": [793], "dimension": [795], "harmonic": [798], "kernel.": [799], "interpretation": [801, 898], "analogy,": [807], "must": [809, 1043], "beclearly": [810], "stated": [811], "such.": [813], "2.": [814], "today": [818], "(1)": [819, 1020], "A": [820, 1021], "computational": [821, 852], "reader": [822], "Immediately": [823], "Using": [825], "numerical": [827, 854], "tools,": [828, 1192], "can:": [830], "construct": [831], "Laplacians": [834], "FEM),": [836], "compute": [837], "eigenvalues,": [839], "estimate": [840], "gaps,": [842], "project": [843], "low-frequency": [846], "components.": [847], "routine": [850], "topology,": [853], "PDEs,": [855], "lattice": [857], "gaugetheory.": [858], "(2)": [859, 1058], "Cross-theory": [860], "classification": [861, 1177], "diagnostics": [866], "spectra),": [873], "(lattice": [875], "simulations": [876], "diagnostics),": [880], "geometric": [883, 1041, 1230], "flows": [884], "(spectral": [885], "collapse).": [886], "Super\u2013Hodge": [888, 1001, 1125], "thus": [889], "acts": [890], "unified": [893], "tool.": [895], "(3)": [896, 1077], "Physical": [897], "light": [900, 907], "emission": [901], "Partially": [902], "identification": [905], "critical": [910], "channel": [912], "with:": [915], "radiation": [918], "conditions,": [919], "linear": [920], "response": [921], "theory.": [922], "However,": [923], "conversion": [925, 949], "holding": [927], "modes": [928, 931, 1288], "lightlike": [930, 1332], "anexplicit": [933], "definition": [934], "structural": [937, 959, 1183], "G\\mathcal": [939, 971], "GG.": [940, 972], "3.": [941], "currently": [945], "provable": [946], "Holding-to-light": [948, 1110], "Open": [950], "implication": [952], "TH\u2192GTLT_H": [953], "\\xrightarrow{\\mathcal": [954], "G}": [955], "T_LTHGTL": [956], "hypothesis,": [960], "theorem.": [963], "Proving": [964], "explicit,coercive": [968], "description": [969], "Universal": [974], "nonlinear": [975, 982], "control": [976], "Out": [977], "reach": [979], "Controlling": [980], "interaction": [983], "term": [984], "\u27e8P\u2264\u03b5N(T),": [985], "T\u2264\u03b5\u27e9\\langle": [986], "P_{\\le\\varepsilon}\\mathcal": [987], "N(T),\\,T_{\\le\\varepsilon}\\rangle\u27e8P\u2264\u03b5N(T),T\u2264\u03b5\u27e9": [988], "uniformly": [989], "precisely": [991], "unresolved": [993], "core": [994], "Yang\u2013Millsproblems.": [999], "isolates": [1003], "difficulty": [1005], "solve": [1009], "it.": [1010], "4.": [1011], "realistically": [1014], "obtained": [1016], "reduction": [1023], "theorem": [1024, 1032], "Realistic": [1025, 1062], "plausible": [1028], "prove": [1030], "form:": [1035], "Any": [1036], "blow-up,": [1037], "mass-gap": [1038], "failure,": [1039], "degeneration": [1042], "manifest": [1044], "asaccumulation": [1045], "quasi-harmonic": [1048], "sector": [1049], "TQT_QTQ.": [1050], "would": [1052], "nontrivial,": [1054], "rigorous,": [1055], "publishable.": [1057], "New": [1059], "invariants": [1061, 1069], "tuple": [1064], "(\u0398H,\u0398Q,\u0398L,gap,b)(\\Theta_H,\\Theta_Q,\\Theta_L,\\mathrm{gap},b)(\u0398H,\u0398Q,\u0398L,gap,b)": [1065], "defines": [1066], "new,": [1067], "computable": [1068], "classify": [1071], "dynamical": [1072], "regimes": [1073], "withoutsolving": [1074], "equations.": [1076], "Conceptual": [1078], "Strong": [1080], "unifies": [1084], "multiple": [1085], "major": [1086, 1145], "theories": [1087], "levelwithout": [1091], "altering": [1092], "their": [1093], "equations": [1094], "definitions.": [1096], "5.": [1097], "Final": [1098], "verdict": [1099], "compatibility:": [1101], "Reader": [1103], "/": [1104, 1184, 1229, 1233, 1235, 1331], "Super\u2013Hodge:": [1105], "implementable": [1106, 1134], "Structural": [1107], "unification:": [1108], "sound": [1109], "emission:": [1111], "hypothesis": [1112], "Clay": [1113, 1161], "resolution:": [1114], "yet": [1116], "maybe": [1118], "verified": [1119], "surclassing": [1121], "One-line": [1122], "conclusion": [1123], "consistent,": [1129], "withexisting": [1131], "computationally": [1133], "tool,": [1138], "andidentifies": [1139], "common": [1141, 1200], "obstruction": [1142, 1305], "underlying": [1143], "several": [1144], "open": [1146], "problems,without": [1147], "claiming": [1148], "resolve": [1150], "them.": [1151], "Version": [1152], "2": [1153, 1257], "Step": [1154, 1195, 1256, 1309], "initial": [1158], "problem": [1159, 1208], "I": [1160], "problems": [1162, 1249], "appear": [1163], "unrelated:": [1164], "Navier\u2013Stokes:": [1165], "blow-up": [1166], "vs": [1167], "regularity": [1168], "Yang\u2013Mills:": [1169], "existence": [1170], "Poincar\u00e9:": [1175], "3-manifolds": [1179], "Hodge,": [1180], "RH,": [1181], "BSD:": [1182], "arithmetic": [1185], "statements": [1186], "Each": [1187], "obstacles.": [1194], "Identify": [1198], "We": [1202, 1263, 1316], "start": [1203], "recognizing": [1205], "governed": [1210], "Laplacian-type": [1213], "operator:": [1214, 1320], "vorticity": [1220], "gauge-covariant": [1223], "Lichnerowicz": [1228], "abstract": [1238], "establishes": [1242], "shared": [1244], "backbone.": [1246], "Outcome:": [1247, 1302], "live": [1250], "stage.": [1255], "Introduce": [1259], "refine": [1264], "classical": [1265], "configuration": [1271], "TTT": [1272], "into:": [1273], "Holding": [1274], "(harmonic)": [1275], "THT_HTH": [1276], "Quasi-harmonic": [1282], "(drift)": [1283], "TQT_QTQ": [1284], "near-kernel,": [1286], "slow": [1287], "Dissipative": [1289], "TDT_DTD": [1290], "high-frequency,": [1292], "fast": [1293], "every": [1303], "localized": [1307], "spectrally.": [1308], "3": [1310], "Add": [1312], "(light": [1314], "channel)": [1315], "introduce": [1317], "\u25a1\u2219=\u2202t2\u2212c2\u0394\u2219\\Box_\\bullet": [1321], "c^2": [1325], "\\Delta_\\bullet\u25a1\u2219=\u2202t2\u2212c2\u0394\u2219": [1326], "define": [1328], "sector:": [1333], "TL=ker\u2061\u25a1\u2219T_L": [1334], "\\ker": [1336], "\\Box_\\bulletTL=ker\u25a1\u2219": [1337], "separates:": [1339], "decay-drive": [1340]}, "primary_topic": {"id": "https://openalex.org/T11022", "display_name": "Spectral Theory in Mathematical Physics", "score": 0.6187000274658203, "subfield": {"id": "https://openalex.org/subfields/2610", "display_name": "Mathematical Physics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11022", "display_name": "Spectral Theory in Mathematical Physics", "score": 0.6187000274658203, "subfield": {"id": "https://openalex.org/subfields/2610", "display_name": "Mathematical Physics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10229", "display_name": "Geometric Analysis and Curvature Flows", "score": 0.11670000106096268, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12536", "display_name": "Topological and Geometric Data Analysis", "score": 0.08219999819993973, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5281/zenodo.18057869", "is_oa": true, "landing_page_url": "https://doi.org/10.5281/zenodo.18057869", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400562", "display_name": "Zenodo (CERN European Organization for Nuclear Research)", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I67311998", "host_organization_name": "European Organization for Nuclear Research", "host_organization_lineage": ["https://openalex.org/I67311998"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "article-journal"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414801144", "doi": "https://doi.org/10.48550/arxiv.2505.14999", "title": "Learning to Rank Chain-of-Thought: Using a Small Model", "publication_year": 2025, "publication_date": "2025-05-21", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5111089095", "display_name": "Eric Hanchen Jiang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Jiang, Eric Hanchen", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5085445569", "display_name": "Haozheng Luo", "orcid": "https://orcid.org/0000-0001-9826-2874", "observed_orcids": ["https://orcid.org/0000-0001-9826-2874"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Luo, Haozheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Pang, Shengyuan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Pang, Shengyuan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5079741060", "display_name": "X. Li", "orcid": "https://orcid.org/0000-0003-1569-0107", "observed_orcids": ["https://orcid.org/0000-0003-1569-0107"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Xiaomin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5055702764", "display_name": "Zhenting Qi", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Qi, Zhenting", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5075978646", "display_name": "Hengli Li", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Hengli", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5103056761", "display_name": "Cheng-Fu Yang", "orcid": "https://orcid.org/0000-0002-6916-2142", "observed_orcids": ["https://orcid.org/0000-0002-6916-2142"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yang, Cheng-Fu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5109734902", "display_name": "Zongyu Lin", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Lin, Zongyu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5044549752", "display_name": "Xinfeng Li", "orcid": "https://orcid.org/0000-0002-9686-4369", "observed_orcids": ["https://orcid.org/0000-0002-9686-4369"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Xinfeng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100641497", "display_name": "Hao Xu", "orcid": "https://orcid.org/0000-0002-4779-9353", "observed_orcids": ["https://orcid.org/0000-0002-4779-9353", "https://orcid.org/0009-0005-4390-4354"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Hao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5087096372", "display_name": "Kai-Wei Chang", "orcid": "https://orcid.org/0000-0001-5365-0072", "observed_orcids": ["https://orcid.org/0000-0001-5365-0072"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chang, Kai-Wei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5059211394", "display_name": "Ying Wu", "orcid": "https://orcid.org/0000-0001-6581-9816", "observed_orcids": ["https://orcid.org/0000-0001-6581-9816"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wu, Ying Nian", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "struggle": [4], "with": [5, 150], "reliable": [6], "mathematical": [7], "reasoning,": [8], "and": [9, 90, 136], "current": [10], "verification": [11], "methods": [12], "are": [13], "often": [14], "computationally": [15], "expensive.": [16], "This": [17, 94, 147], "paper": [18], "introduces": [19], "the": [20, 61, 80, 101, 116], "Energy": [21], "Outcome": [22], "Reward": [23], "Model": [24], "(EORM),": [25], "a": [26, 106, 156], "highly": [27], "efficient,": [28], "lightweight": [29], "post-hoc": [30], "verifier": [31], "designed": [32], "to": [33, 42, 48, 86, 112, 133], "address": [34], "this": [35], "challenge.": [36], "EORM": [37, 78, 130, 154], "uses": [38], "an": [39], "energy-based": [40], "framework": [41], "rank": [43], "Chain-of-Thought": [44], "(CoT)": [45], "solutions,": [46], "learning": [47], "distinguish": [49], "correct": [50], "from": [51, 105], "incorrect": [52], "reasoning": [53, 103], "using": [54], "only": [55, 67], "simple": [56], "outcome": [57], "labels,": [58], "thus": [59], "eliminating": [60], "need": [62], "for": [63, 159], "expensive": [64], "annotations.": [65], "With": [66], "55M": [68], "parameters,": [69], "over": [70], "127": [71], "times": [72], "smaller": [73], "than": [74], "typical": [75], "reward": [76], "models,": [77, 138], "boosts": [79], "accuracy": [81, 117], "of": [82, 108, 118, 144], "Llama": [83], "3": [84], "8B": [85], "90.7\\%": [87], "on": [88, 92], "GSM8k": [89], "63.7\\%": [91], "MATH.": [93], "performance": [95], "is": [96], "achieved": [97], "by": [98], "efficiently": [99], "selecting": [100], "optimal": [102], "path": [104], "pool": [107], "candidates,": [109], "allowing": [110], "it": [111, 140], "match": [113], "or": [114], "exceed": [115], "far": [119], "more": [120, 161], "resource-intensive": [121], "Best-of-N": [122], "sampling": [123], "techniques.": [124], "Crucially,": [125], "our": [126], "experiments": [127], "show": [128], "that": [129], "generalizes": [131], "effectively": [132], "out-of-distribution": [134], "problems": [135], "unseen": [137], "indicating": [139], "learns": [141], "fundamental": [142], "principles": [143], "valid": [145], "reasoning.": [146], "robustness,": [148], "combined": [149], "its": [151], "efficiency,": [152], "establishes": [153], "as": [155], "practical": [157], "tool": [158], "deploying": [160], "dependable": [162], "LLMs": [163], "in": [164], "complex,": [165], "real-world": [166], "applications.": [167]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5220000147819519, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5220000147819519, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.11050000041723251, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11689", "display_name": "Adversarial Robustness in Machine Learning", "score": 0.0414000004529953, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2505.14999", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2505.14999", "pdf_url": "https://arxiv.org/pdf/2505.14999", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417357116", "doi": "https://doi.org/10.48550/arxiv.2506.10947", "title": "Spurious Rewards: Rethinking Training Signals in RLVR", "publication_year": 2025, "publication_date": "2025-06-12", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5047742348", "display_name": "Rulin Shao", "orcid": "https://orcid.org/0000-0002-8228-2106", "observed_orcids": ["https://orcid.org/0000-0002-8228-2106"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Shao, Rulin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5022934996", "display_name": "Shuyue Stella Li", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Shuyue Stella", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101888897", "display_name": "Rui Xin", "orcid": "https://orcid.org/0000-0002-0224-3431", "observed_orcids": ["https://orcid.org/0000-0002-0224-3431"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xin, Rui", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5005776608", "display_name": "Scott Geng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Geng, Scott", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5113844261", "display_name": "Yiping Wang", "orcid": "https://orcid.org/0000-0002-7540-6449", "observed_orcids": ["https://orcid.org/0000-0002-7540-6449"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Yiping", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5028243041", "display_name": "Sewoong Oh", "orcid": "https://orcid.org/0000-0002-8975-8306", "observed_orcids": ["https://orcid.org/0000-0002-8975-8306"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Oh, Sewoong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5033061754", "display_name": "Simon S. Du", "orcid": "https://orcid.org/0000-0003-0056-8299", "observed_orcids": ["https://orcid.org/0000-0003-0056-8299"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Du, Simon Shaolei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5083548887", "display_name": "Nathan Lambert", "orcid": "https://orcid.org/0000-0002-9997-6817", "observed_orcids": ["https://orcid.org/0000-0002-9997-6817"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Lambert, Nathan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5039158419", "display_name": "Sewon Min", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Min, Sewon", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5032451496", "display_name": "Ranjay Krishna", "orcid": "https://orcid.org/0000-0001-8784-2531", "observed_orcids": ["https://orcid.org/0000-0001-8784-2531"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Krishna, Ranjay", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5062910836", "display_name": "Yulia Tsvetkov", "orcid": "https://orcid.org/0000-0002-4634-7128", "observed_orcids": ["https://orcid.org/0000-0002-4634-7128"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Tsvetkov, Yulia", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5082305994", "display_name": "Hannaneh Hajishirzi", "orcid": "https://orcid.org/0000-0002-1055-6657", "observed_orcids": ["https://orcid.org/0000-0002-1055-6657"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Hajishirzi, Hannaneh", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5079904764", "display_name": "Pang Wei Koh", "orcid": "https://orcid.org/0000-0003-4330-6969", "observed_orcids": ["https://orcid.org/0000-0003-4330-6969"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Koh, Pang Wei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5067919401", "display_name": "Luke Zettlemoyer", "orcid": "https://orcid.org/0009-0008-8296-0764", "observed_orcids": ["https://orcid.org/0009-0008-8296-0764"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zettlemoyer, Luke", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"We": [0], "show": [1, 67], "that": [2, 22, 68, 142, 194], "reinforcement": [3], "learning": [4], "with": [5, 19, 30, 38, 125], "verifiable": [6], "rewards": [7, 21, 141, 193], "(RLVR)": [8], "can": [9, 79, 185], "elicit": [10], "strong": [11], "mathematical": [12], "reasoning": [13, 106, 108], "in": [14, 99, 109], "certain": [15], "language": [16], "models": [17, 147, 173, 189], "even": [18, 27, 86, 190], "spurious": [20, 126, 140], "have": [23], "little,": [24], "no,": [25], "or": [26, 160], "negative": [28], "correlation": [29], "the": [31, 55, 75, 129, 165], "correct": [32], "answer.": [33], "For": [34], "example,": [35], "RLVR": [36], "training": [37], "GRPO": [39, 69], "improves": [40], "MATH-500": [41], "performance": [42], "for": [43, 145, 153], "Qwen2.5-Math-7B": [44], "by": [45], "21.4": [46], "percentage": [47], "points": [48], "using": [49], "randomly": [50], "assigned": [51], "rewards,": [52], "nearly": [53], "matching": [54], "29.1-point": [56], "gain": [57], "from": [58, 74, 118, 191], "ground-truth": [59], "rewards.": [60, 89, 127], "To": [61], "explain": [62], "this": [63], "counterintuitive": [64], "observation,": [65], "we": [66, 94, 103], "exhibits": [70], "a": [71, 91, 178], "clipping": [72], "bias": [73], "clip": [76], "term,": [77], "which": [78, 102], "amplify": [80], "high-prior": [81], "behaviors": [82, 134], "learned": [83], "during": [84], "pretraining": [85], "without": [87, 111], "informative": [88], "As": [90], "case": [92], "study,": [93], "identify": [95], "one": [96], "such": [97, 132, 157], "behavior": [98], "Qwen2.5-Math": [100], "models,": [101], "call": [104], "code": [105, 110, 113], "--": [107], "actual": [112], "execution;": [114], "code-reasoning": [115], "frequency": [116], "increases": [117], "65": [119], "percent": [120, 124], "to": [121, 150], "over": [122], "90": [123], "However,": [128], "presence": [130], "of": [131, 167], "amplifiable": [133], "is": [135], "highly": [136], "model-dependent.": [137], "In": [138], "practice,": [139], "are": [143], "effective": [144], "Qwen": [146, 188], "often": [148], "fail": [149], "produce": [151], "gains": [152, 184], "other": [154], "model": [155], "families,": [156], "as": [158], "Llama3": [159], "OLMo2.": [161], "Our": [162], "results": [163], "highlight": [164], "importance": [166], "validating": [168], "RL": [169], "methods": [170], "across": [171], "diverse": [172], "rather": [174], "than": [175], "relying": [176], "on": [177, 187], "single": [179], "de": [180], "facto": [181], "choice:": [182], "large": [183], "arise": [186], "random": [192], "do": [195], "not": [196], "reflect": [197], "genuine": [198], "capability": [199], "improvements.": [200]}, "primary_topic": {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.4099000096321106, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.4099000096321106, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.11789999902248383, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.04749999940395355, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2506.10947", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2506.10947", "pdf_url": "https://arxiv.org/pdf/2506.10947", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415318102", "doi": "https://doi.org/10.48550/arxiv.2510.07169", "title": "More Data or Better Data? A Critical Analysis of Data Selection and Synthesis for Mathematical Reasoning", "publication_year": 2025, "publication_date": "2025-10-08", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Zhao, Yike", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I66867065", "display_name": "East China Normal University", "ror": "https://ror.org/02n96ep67", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I66867065"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Zhao, Yike", "raw_affiliation_strings": ["East China Normal University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "East China Normal University", "institution_ids": ["https://openalex.org/I66867065"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101580861", "display_name": "Simin Guo", "orcid": "https://orcid.org/0000-0002-5729-616X", "observed_orcids": ["https://orcid.org/0000-0002-5729-616X"]}, "institutions": [{"id": "https://openalex.org/I40347166", "display_name": "University of Chicago", "ror": "https://ror.org/024mw5h28", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I40347166"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Guo, Simin", "raw_affiliation_strings": ["University of Chicago"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University of Chicago", "institution_ids": ["https://openalex.org/I40347166"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5001241589", "display_name": "Ziqing Yang", "orcid": "https://orcid.org/0000-0002-0492-0862", "observed_orcids": ["https://orcid.org/0000-0002-0492-0862", "https://orcid.org/0000-0001-8823-575X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yang, Ziqing", "raw_affiliation_strings": ["Independent Researcher"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Independent Researcher", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5018437726", "display_name": "Shifan Han", "orcid": "https://orcid.org/0000-0003-4448-1689", "observed_orcids": ["https://orcid.org/0000-0003-4448-1689"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Han, Shifan", "raw_affiliation_strings": ["Independent Researcher"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Independent Researcher", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5010087030", "display_name": "Dahua Lin", "orcid": "https://orcid.org/0000-0002-8865-7896", "observed_orcids": ["https://orcid.org/0000-0002-8865-7896"]}, "institutions": [{"id": "https://openalex.org/I177725633", "display_name": "Chinese University of Hong Kong", "ror": "https://ror.org/00t33hh48", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I177725633"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Lin, Dahua", "raw_affiliation_strings": ["The Chinese University of Hong Kong"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "The Chinese University of Hong Kong", "institution_ids": ["https://openalex.org/I177725633"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100740044", "display_name": "Fei Zhi Tan", "orcid": "https://orcid.org/0000-0003-2423-631X", "observed_orcids": ["https://orcid.org/0000-0003-2423-631X"]}, "institutions": [{"id": "https://openalex.org/I66867065", "display_name": "East China Normal University", "ror": "https://ror.org/02n96ep67", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I66867065"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Tan, Fei", "raw_affiliation_strings": ["East China Normal University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "East China Normal University", "institution_ids": ["https://openalex.org/I66867065"]}]}], "abstract_inverted_index": {"The": [0], "reasoning": [1, 148], "capabilities": [2], "of": [3, 22, 47], "Large": [4], "Language": [5], "Models": [6], "(LLMs)": [7], "play": [8], "a": [9, 44, 60], "critical": [10], "role": [11], "in": [12, 34, 91], "many": [13], "downstream": [14], "tasks,": [15], "yet": [16], "depend": [17], "strongly": [18], "on": [19, 137], "the": [20], "quality": [21], "training": [23, 66, 114], "data.": [24], "Despite": [25], "various": [26], "proposed": [27], "data": [28, 51, 74, 90, 105, 115, 123], "construction": [29], "methods,": [30], "their": [31], "practical": [32, 79], "utility": [33], "real-world": [35, 147], "pipelines": [36], "remains": [37], "underexplored.": [38], "In": [39], "this": [40, 131], "work,": [41], "we": [42], "conduct": [43], "comprehensive": [45], "analysis": [46], "open-source": [48], "datasets": [49], "and": [50, 67, 77, 125], "synthesis": [52], "techniques": [53], "for": [54, 82, 112, 146], "mathematical": [55], "reasoning,": [56], "evaluating": [57], "them": [58], "under": [59], "unified": [61], "pipeline": [62], "designed": [63], "to": [64, 116, 139], "mirror": [65], "deployment": [68], "scenarios.": [69], "We": [70, 129], "further": [71, 135], "distill": [72], "effective": [73], "selection": [75], "strategies": [76], "identify": [78], "methods": [80], "suitable": [81], "industrial": [83], "applications.": [84], "Our": [85], "findings": [86], "highlight": [87], "that": [88], "structuring": [89], "more": [92], "interpretable": [93], "formats,": [94], "or": [95], "distilling": [96], "from": [97], "stronger": [98], "models": [99], "often": [100], "outweighs": [101], "simply": [102], "scaling": [103], "up": [104], "volume.": [106], "This": [107], "study": [108], "provides": [109], "actionable": [110], "guidance": [111], "integrating": [113], "enhance": [117], "LLM": [118], "capabilities,": [119], "supporting": [120], "both": [121], "cost-effective": [122], "curation": [124], "scalable": [126], "model": [127], "enhancement.": [128], "hope": [130], "work": [132], "will": [133], "inspire": [134], "research": [136], "how": [138], "balance": [140], "\"more": [141], "data\"": [142, 145], "versus": [143], "\"better": [144], "tasks.": [149]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8788999915122986, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8788999915122986, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.026100000366568565, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.020600000396370888, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.07169", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.07169", "pdf_url": "https://arxiv.org/pdf/2510.07169", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417104745", "doi": "https://doi.org/10.48550/arxiv.2510.11718", "title": "CodePlot-CoT: Mathematical Visual Reasoning by Thinking with Code-Driven Images", "publication_year": 2025, "publication_date": "2025-10-13", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5054320268", "display_name": "Chengqi Duan", "orcid": "https://orcid.org/0009-0004-5737-019X", "observed_orcids": ["https://orcid.org/0009-0004-5737-019X"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Duan, Chengqi", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101853880", "display_name": "Kaiyue Sun", "orcid": "https://orcid.org/0000-0003-1401-4051", "observed_orcids": ["https://orcid.org/0000-0003-1401-4051"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Sun, Kaiyue", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5025456709", "display_name": "Rongyao Fang", "orcid": "https://orcid.org/0000-0002-8010-4808", "observed_orcids": ["https://orcid.org/0000-0002-8010-4808"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Fang, Rongyao", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5012273415", "display_name": "Manyuan Zhang", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Zhang, Manyuan", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5061181883", "display_name": "Feng Yan", "orcid": "https://orcid.org/0000-0003-1841-0318", "observed_orcids": ["https://orcid.org/0000-0003-1841-0318"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Feng, Yan", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5083405649", "display_name": "Ying Luo", "orcid": "https://orcid.org/0000-0003-3074-8813", "observed_orcids": ["https://orcid.org/0000-0003-3074-8813", "https://orcid.org/0000-0003-1460-4289"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Luo, Ying", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Liu, Yufang", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Liu, Yufang", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101645620", "display_name": "Ke Wang", "orcid": "https://orcid.org/0000-0001-5783-1655", "observed_orcids": ["https://orcid.org/0000-0001-5783-1655"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Wang, Ke", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Pei, Peng", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Pei, Peng", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5068562468", "display_name": "Xunliang Cai", "orcid": "https://orcid.org/0000-0003-0685-9556", "observed_orcids": ["https://orcid.org/0000-0003-0685-9556"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Cai, Xunliang", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100732450", "display_name": "Hongsheng Li", "orcid": "https://orcid.org/0000-0002-2664-7975", "observed_orcids": ["https://orcid.org/0000-0002-2664-7975"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Li, Hongsheng", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Ma, Yi", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Ma, Yi", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5027234036", "display_name": "Xihui Liu", "orcid": "https://orcid.org/0000-0002-1791-8745", "observed_orcids": ["https://orcid.org/0000-0002-1791-8745", "https://orcid.org/0000-0003-1831-9952"]}, "institutions": [{"id": "https://openalex.org/I889458895", "display_name": "University of Hong Kong", "ror": "https://ror.org/02zhqgq86", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I889458895"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Liu, Xihui", "raw_affiliation_strings": ["CUHK LLM/VLM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "CUHK LLM/VLM", "institution_ids": ["https://openalex.org/I889458895"]}]}], "abstract_inverted_index": {"Recent": [0], "advances": [1], "in": [2, 16, 87], "Large": [3], "Language": [4, 9], "Models": [5, 10], "(LLMs)": [6], "and": [7, 45, 62, 68, 129, 212, 223, 238], "Vision": [8], "(VLMs)": [11], "have": [12], "shown": [13], "significant": [14], "progress": [15], "mathematical": [17, 115, 156, 172, 210], "reasoning,": [18], "yet": [19], "they": [20], "still": [21], "face": [22], "a": [23, 79, 148, 205], "critical": [24], "bottleneck": [25], "with": [26, 85, 134, 216], "problems": [27, 133], "requiring": [28], "visual": [29], "assistance,": [30], "such": [31, 71, 227], "as": [32, 98, 100, 110], "drawing": [33], "auxiliary": [34], "lines": [35], "or": [36], "plotting": [37, 102], "functions": [38], "to": [39, 49, 94, 113, 141, 182], "solve": [40, 114], "the": [41, 65, 92, 124, 167, 194, 214, 217], "problems.": [42, 116, 173, 228], "Most": [43], "LLMs": [44], "VLMs": [46], "are": [47], "constrained": [48], "text-only": [50], "reasoning": [51, 97, 200, 211], "chains,": [52], "while": [53], "multimodal": [54, 209], "unified": [55], "models": [56, 240], "that": [57, 177], "can": [58], "generate": [59, 95], "interleaved": [60], "text": [61, 96], "images": [63, 109], "lack": [64], "necessary": [66], "precision": [67], "controllability": [69], "for": [70, 83, 131, 153, 170, 208, 226], "tasks.": [72], "To": [73, 117, 229], "address": [74], "this,": [75, 119], "we": [76, 120, 146, 165, 233], "propose": [77], "CodePlot-CoT,": [78], "code-driven": [80, 199], "Chain-of-Thought": [81], "paradigm": [82], "\"thinking": [84], "images\"": [86], "mathematics.": [88], "Our": [89, 202], "approach": [90, 225], "leverages": [91], "VLM": [93], "well": [99], "executable": [101], "code,": [103, 237], "which": [104], "is": [105], "then": [106], "rendered": [107], "into": [108, 158], "\"visual": [111], "thought\",": [112], "achieve": [118], "first": [121, 125, 218], "construct": [122], "Math-VR,": [123], "large-scale,": [126], "bilingual": [127], "dataset": [128], "benchmark": [130], "Mathematics": [132], "Visual": [135], "Reasoning,": [136], "comprising": [137], "178K": [138], "samples.": [139], "Second,": [140], "create": [142], "high-quality": [143], "training": [144, 163], "data,": [145, 164], "develop": [147], "state-of-the-art": [149], "image-to-code": [150], "converter": [151], "specialized": [152], "parsing": [154], "complex": [155], "figures": [157], "codes.": [159], "Finally,": [160], "using": [161], "these": [162], "train": [166], "CodePlot-CoT": [168], "model": [169, 179, 187], "solving": [171], "Experimental": [174], "results": [175], "show": [176], "our": [178, 189, 197, 235], "achieves": [180], "up": [181], "21%": [183], "increase": [184], "over": [185], "base": [186], "on": [188], "new": [190, 206], "benchmark,": [191, 222], "fully": [192], "validating": [193], "efficacy": [195], "of": [196], "proposed": [198], "paradigm.": [201], "work": [203], "opens": [204], "direction": [207], "provides": [213], "community": [215], "large-scale": [219], "dataset,": [220], "comprehensive": [221], "strong": [224], "facilitate": [230], "future": [231], "research,": [232], "make": [234], "datasets,": [236], "pretrained": [239], "publicly": [241], "available": [242], "at": [243], "https://github.com/HKU-MMLab/Math-VR-CodePlot-CoT.": [244]}, "primary_topic": {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.9589999914169312, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.9589999914169312, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.01549999974668026, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.007499999832361937, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.11718", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.11718", "pdf_url": "https://arxiv.org/pdf/2510.11718", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by-nc-sa", "license_id": "https://openalex.org/licenses/cc-by-nc-sa", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4410145882", "doi": "https://doi.org/10.1007/s44366-025-0053-z", "title": "MathEval: A Comprehensive Benchmark for Evaluating Large Language Models on Mathematical Reasoning Capabilities", "publication_year": 2025, "publication_date": "2025-05-07", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5078859530", "display_name": "Tianqiao Liu", "orcid": "https://orcid.org/0000-0001-7997-3232", "observed_orcids": ["https://orcid.org/0000-0001-7997-3232", "https://orcid.org/0000-0002-6287-207X"]}, "institutions": [{"id": "https://openalex.org/I159948400", "display_name": "Jinan University", "ror": "https://ror.org/02xe5ns62", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I159948400"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Tianqiao Liu", "raw_affiliation_strings": ["Guangdong Institute of Smart Education, Jinan University, Guangzhou, 510632, China", "TAL Education Group, Beijing, 102206, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Guangdong Institute of Smart Education, Jinan University, Guangzhou, 510632, China", "institution_ids": ["https://openalex.org/I159948400"]}, {"raw_affiliation_string": "TAL Education Group, Beijing, 102206, China", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5022889358", "display_name": "Zui Chen", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zui Chen", "raw_affiliation_strings": ["TAL Education Group, Beijing, 102206, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TAL Education Group, Beijing, 102206, China", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Zhensheng Fang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhensheng Fang", "raw_affiliation_strings": ["TAL Education Group, Beijing, 102206, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TAL Education Group, Beijing, 102206, China", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102785655", "display_name": "Weiqi Luo", "orcid": "https://orcid.org/0000-0002-8999-6064", "observed_orcids": ["https://orcid.org/0000-0002-8999-6064"]}, "institutions": [{"id": "https://openalex.org/I159948400", "display_name": "Jinan University", "ror": "https://ror.org/02xe5ns62", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I159948400"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Weiqi Luo", "raw_affiliation_strings": ["Guangdong Institute of Smart Education, Jinan University, Guangzhou, 510632, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Guangdong Institute of Smart Education, Jinan University, Guangzhou, 510632, China", "institution_ids": ["https://openalex.org/I159948400"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Mi Tian", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Mi Tian", "raw_affiliation_strings": ["TAL Education Group, Beijing, 102206, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TAL Education Group, Beijing, 102206, China", "institution_ids": []}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5102799101", "display_name": "Zitao Liu", "orcid": "https://orcid.org/0000-0001-7600-6079", "observed_orcids": ["https://orcid.org/0000-0001-7600-6079"]}, "institutions": [{"id": "https://openalex.org/I159948400", "display_name": "Jinan University", "ror": "https://ror.org/02xe5ns62", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I159948400"]}], "countries": ["CN"], "is_corresponding": true, "raw_author_name": "Zitao Liu", "raw_affiliation_strings": ["Guangdong Institute of Smart Education, Jinan University, Guangzhou, 510632, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Guangdong Institute of Smart Education, Jinan University, Guangzhou, 510632, China", "institution_ids": ["https://openalex.org/I159948400"]}]}], "abstract_inverted_index": null, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7113999724388123, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7113999724388123, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.06790000200271606, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.05249999836087227, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.1007/s44366-025-0053-z", "is_oa": false, "landing_page_url": "https://doi.org/10.1007/s44366-025-0053-z", "pdf_url": null, "source": {"id": "https://openalex.org/S4404674969", "display_name": "Frontiers of digital education.", "issn_l": "2097-3918", "issn": ["2097-3918", "2097-3926"], "is_oa": false, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/P4310311889", "host_organization_name": "Higher Education Press", "host_organization_lineage": ["https://openalex.org/P4310311889"], "host_organization_lineage_names": ["Higher Education Press"], "type": "journal"}, "license": null, "license_id": null, "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Frontiers of Digital Education", "raw_type": "journal-article"}, "cited_by_count": 5, "is_retracted": false}, {"id": "https://openalex.org/W4417045551", "doi": "https://doi.org/10.2139/ssrn.5869482", "title": "Endogenous Memory for Large Language Models: An Adaptive Framework from Data Streams to Parameterization", "publication_year": 2025, "publication_date": "2025-01-01", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5120714846", "display_name": "Jiacheng Wei", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Jiacheng Wei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "yanxia chang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "yanxia chang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "xiao zhang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "xiao zhang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5026806196", "display_name": "Dongzhuo Li", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "dongrun li", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5101469134", "display_name": "Fei Ling Wu", "orcid": "https://orcid.org/0000-0002-2240-4829", "observed_orcids": ["https://orcid.org/0000-0002-2240-4829"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "faguo wu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": null, "primary_topic": {"id": "https://openalex.org/T12761", "display_name": "Data Stream Mining Techniques", "score": 0.5512999892234802, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T12761", "display_name": "Data Stream Mining Techniques", "score": 0.5512999892234802, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.2070000022649765, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.18479999899864197, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.2139/ssrn.5869482", "is_oa": true, "landing_page_url": "https://doi.org/10.2139/ssrn.5869482", "pdf_url": null, "source": {"id": "https://openalex.org/S4210172589", "display_name": "SSRN Electronic Journal", "issn_l": "1556-5068", "issn": ["1556-5068"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I1318003438", "host_organization_name": "RELX Group (Netherlands)", "host_organization_lineage": ["https://openalex.org/I1318003438"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "acceptedVersion", "is_accepted": true, "is_published": false, "raw_source_name": null, "raw_type": "posted-content"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4406317020", "doi": "https://doi.org/10.48550/arxiv.2501.05765", "title": "Deontic Temporal Logic for Formal Verification of AI Ethics", "publication_year": 2025, "publication_date": "2025-01-10", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5108937880", "display_name": "Tanu Priya", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "V., Priya T.", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5055571512", "display_name": "Shrisha Rao", "orcid": "https://orcid.org/0000-0003-0625-5103", "observed_orcids": ["https://orcid.org/0000-0003-0625-5103", "https://orcid.org/0000-0001-5559-7287"]}, "institutions": [{"id": "https://openalex.org/I181514455", "display_name": "International Institute of Information Technology Bangalore", "ror": "https://ror.org/05h9eqy10", "country_code": "IN", "type": "education", "lineage": ["https://openalex.org/I181514455"]}], "countries": ["IN"], "is_corresponding": false, "raw_author_name": "Rao, Shrisha", "raw_affiliation_strings": ["International Institute of Information Technology - Bangalore"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "International Institute of Information Technology - Bangalore", "institution_ids": ["https://openalex.org/I181514455"]}]}], "abstract_inverted_index": {"Ensuring": [0], "ethical": [1, 39, 58, 79, 95, 124, 168, 185], "behavior": [2, 40, 59, 96], "in": [3, 26, 182, 187], "Artificial": [4], "Intelligence": [5], "(AI)": [6], "systems": [7, 99, 132, 151, 162], "amidst": [8], "their": [9], "increasing": [10], "ubiquity": [11], "and": [12, 36, 55, 75, 84, 118, 129, 173], "influence": [13], "is": [14, 29], "a": [15, 30, 47], "major": [16], "concern": [17], "the": [18, 38, 57, 94, 105, 112, 115, 127, 140, 153, 176, 179], "world": [19], "over.": [20], "The": [21, 86, 102, 156], "use": [22, 141], "of": [23, 41, 60, 97, 107, 114, 126, 142, 178], "formal": [24, 157], "methods": [25], "AI": [27, 42, 61, 98, 121, 189], "ethics": [28, 113], "possible": [31], "crucial": [32], "approach": [33], "for": [34], "specifying": [35], "verifying": [37], "systems.": [43, 122], "This": [44], "paper": [45], "proposes": [46], "formalization": [48, 87, 109, 181], "based": [49], "on": [50, 64], "deontic": [51, 136], "logic": [52], "to": [53, 68, 77, 82, 91, 147, 164, 171], "define": [54], "evaluate": [56, 104], "systems,": [62], "focusing": [63], "system-level": [65], "specifications,": [66], "contributing": [67], "this": [69, 108], "important": [70], "goal.": [71], "It": [72], "introduces": [73], "axioms": [74], "theorems": [76], "capture": [78], "requirements": [80], "related": [81, 170], "fairness": [83, 172], "explainability.": [85], "incorporates": [88], "temporal": [89], "operators": [90], "reason": [92], "about": [93], "over": [100], "time.": [101], "authors": [103], "effectiveness": [106, 177], "by": [110], "assessing": [111], "real-world": [116, 188], "COMPAS": [117, 128], "loan": [119, 130], "prediction": [120, 131], "Various": [123], "properties": [125, 169], "are": [133], "encoded": [134], "using": [135], "logical": [137], "formulas,": [138], "allowing": [139], "an": [143], "automated": [144], "theorem": [145], "prover": [146], "verify": [148], "whether": [149], "these": [150], "satisfy": [152], "defined": [154], "properties.": [155], "verification": [158], "reveals": [159], "that": [160], "both": [161], "fail": [163], "fulfill": [165], "certain": [166], "key": [167], "non-discrimination,": [174], "demonstrating": [175], "proposed": [180], "identifying": [183], "potential": [184], "issues": [186], "applications.": [190]}, "primary_topic": {"id": "https://openalex.org/T10883", "display_name": "Ethics and Social Impacts of AI", "score": 0.951200008392334, "subfield": {"id": "https://openalex.org/subfields/3311", "display_name": "Safety Research"}, "field": {"id": "https://openalex.org/fields/33", "display_name": "Social Sciences"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, "topics": [{"id": "https://openalex.org/T10883", "display_name": "Ethics and Social Impacts of AI", "score": 0.951200008392334, "subfield": {"id": "https://openalex.org/subfields/3311", "display_name": "Safety Research"}, "field": {"id": "https://openalex.org/fields/33", "display_name": "Social Sciences"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T11010", "display_name": "Logic, Reasoning, and Knowledge", "score": 0.03880000114440918, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.0015999999595806003, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2501.05765", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2501.05765", "pdf_url": "https://arxiv.org/pdf/2501.05765", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by-nc-sa", "license_id": "https://openalex.org/licenses/cc-by-nc-sa", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4407373378", "doi": "https://doi.org/10.32620/oikit.2024.102.07", "title": "\u041c\u041e\u0416\u041b\u0418\u0412\u041e\u0421\u0422\u0406 \u0412\u0418\u041a\u041e\u0420\u0418\u0421\u0422\u0410\u041d\u041d\u042f \u0406\u041d\u0422\u0415\u041b\u0415\u041a\u0422\u0423\u0410\u041b\u042c\u041d\u0418\u0425 \u041d\u0410\u0412\u0427\u0410\u041b\u042c\u041d\u0418\u0425 \u0421\u0418\u0421\u0422\u0415\u041c (ITS) \u0412 \u0417\u0412\u041e \u0414\u041b\u042f \u041a\u0423\u0420\u0421\u0406\u0412 \u0412\u0418\u0429\u041e\u0407 \u041c\u0410\u0422\u0415\u041c\u0410\u0422\u0418\u041a\u0418", "publication_year": 2025, "publication_date": "2025-02-10", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5117398645", "display_name": "\u0410. \u0413. \u0427\u0443\u0445\u0440\u0430\u0439", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "\u0410. \u0413. \u0427\u0443\u0445\u0440\u0430\u0439", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5116230656", "display_name": "\u0422. \u041b. \u0421\u0442\u043e\u043b\u044f\u0440\u0435\u043d\u043a\u043e", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "\u0422. \u041b. \u0421\u0442\u043e\u043b\u044f\u0440\u0435\u043d\u043a\u043e", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5116230657", "display_name": "\u041e. \u041e. \u0404\u0432\u0434\u043e\u043a\u0438\u043c\u043e\u0432", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "\u041e. \u041e. \u0404\u0432\u0434\u043e\u043a\u0438\u043c\u043e\u0432", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5116230658", "display_name": "\u0412. \u0410. \u0414\u0435\u043c\u2019\u044f\u043d\u0435\u043d\u043a\u043e", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "\u0412. \u0410. \u0414\u0435\u043c\u2019\u044f\u043d\u0435\u043d\u043a\u043e", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"The": [0, 268], "current": [1], "state": [2], "of": [3, 32, 46, 51, 68, 77, 83, 93, 118, 124, 151, 164, 173, 182, 188, 215, 228, 245, 249, 281], "mathematical": [4, 47, 84, 230, 246, 282], "knowledge": [5], "among": [6], "school": [7], "graduates": [8], "in": [9, 24, 115, 206, 278], "Ukraine": [10], "and": [11, 56, 66, 106, 141, 160, 176, 185, 204, 218, 222, 260, 266, 285], "worldwide": [12], "is": [13, 62, 98, 130, 154, 178, 192, 225, 239], "analyzed.": [14], "A": [15, 91, 144, 180, 220], "negative": [16], "trend": [17], "has": [18], "been": [19], "identified,": [20], "requiring": [21], "urgent": [22], "changes": [23], "educational": [25, 200], "approaches.": [26], "This": [27], "study": [28, 79, 251], "explores": [29], "the": [30, 40, 44, 78, 81, 116, 149, 162, 171, 243, 250, 279], "potential": [31, 163], "using": [33, 165], "intelligent": [34, 87, 257], "computer-based": [35, 88], "learning": [36, 119, 258], "programs": [37, 53], "to": [38, 59, 132, 147, 169, 194, 288], "mitigate": [39], "adverse": [41], "effects": [42], "on": [43, 103, 110], "quality": [45, 244], "education.": [48], "Existing": [49, 202], "models": [50, 85, 120, 191, 197], "such": [52, 137, 152], "are": [54, 126, 209], "analyzed,": [55, 210], "their": [57, 157, 216], "application": [58], "mathematics": [60, 89], "teaching": [61, 148], "examined.": [63, 127], "Mathematical": [64], "modeling": [65], "analysis": [67], "open-source": [69], "information": [70], "were": [71], "used": [72], "as": [73, 138], "research": [74], "methods.The": [75], "results": [76, 248], "include": [80], "development": [82, 227], "for": [86, 121, 198, 233, 241, 255, 264, 275], "learning.": [90], "classification": [92], "tasks": [94, 101, 108, 125, 133, 153, 175], "into": [95], "two": [96], "subgroups": [97], "proposed:": [99], "simple": [100], "(focused": [102, 109], "algorithm": [104, 111, 135], "application)": [105], "complex": [107], "creation).": [112], "Key": [113], "differences": [114], "implementation": [117], "each": [122], "class": [123], "Particular": [128], "attention": [129], "paid": [131], "involving": [134], "creation,": [136], "theorem": [139], "proving": [140], "formula": [142], "derivation.": [143], "novel": [145], "approach": [146], "resolution": [150], "proposed,": [155], "considering": [156], "unique": [158, 174], "structure": [159], "complexity.Additionally,": [161], "generative": [166, 189], "artificial": [167], "intelligence": [168], "automate": [170], "creation": [172], "databases": [177], "explored.": [179], "combination": [181], "algorithmic": [183], "precision": [184], "lexical": [186], "flexibility": [187], "AI": [190], "proposed": [193], "build": [195], "comprehensive": [196], "any": [199], "task.": [201], "projects": [203], "technologies": [205], "this": [207], "field": [208, 280], "with": [211, 272], "a": [212, 229, 273], "comparative": [213], "assessment": [214], "advantages": [217], "disadvantages.": [219], "promising": [221], "underexplored": [223], "area": [224], "identified\u2014the": [226], "user": [231], "interface": [232], "automated": [234], "theorem-proof": [235], "verification": [236], "systems,": [237], "which": [238], "crucial": [240], "improving": [242], "education.The": [247], "open": [252], "new": [253], "opportunities": [254], "enhancing": [256], "systems": [259], "have": [261], "practical": [262], "value": [263], "engineers": [265], "entrepreneurs.": [267], "work": [269], "provides": [270], "researchers": [271], "foundation": [274], "further": [276], "investigations": [277], "education": [283], "automation": [284], "its": [286], "adaptation": [287], "modern": [289], "challenges.": [290]}, "primary_topic": {"id": "https://openalex.org/T14414", "display_name": "Artificial Intelligence in Education", "score": 0.5985999703407288, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T14414", "display_name": "Artificial Intelligence in Education", "score": 0.5985999703407288, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11902", "display_name": "Intelligent Tutoring Systems and Adaptive Learning", "score": 0.34769999980926514, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10130", "display_name": "Mathematics Education and Teaching Techniques", "score": 0.008899999782443047, "subfield": {"id": "https://openalex.org/subfields/3304", "display_name": "Education"}, "field": {"id": "https://openalex.org/fields/33", "display_name": "Social Sciences"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}], "primary_location": {"id": "doi:10.32620/oikit.2024.102.07", "is_oa": true, "landing_page_url": "https://doi.org/10.32620/oikit.2024.102.07", "pdf_url": "http://nti.khai.edu/ojs/index.php/oikit/article/download/2715/2575", "source": {"id": "https://openalex.org/S4210228359", "display_name": "Open Information and Computer Integrated Technologies", "issn_l": "2071-1077", "issn": ["2071-1077", "2663-2411"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/P4310321196", "host_organization_name": "National Aerospace University \u2013 Kharkiv Aviation Institute", "host_organization_lineage": ["https://openalex.org/P4310321196"], "host_organization_lineage_names": ["National Aerospace University \u2013 Kharkiv Aviation Institute"], "type": "journal"}, "license": null, "license_id": null, "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Open Information and Computer Integrated Technologies", "raw_type": "journal-article"}, "cited_by_count": 1, "is_retracted": false}, {"id": "https://openalex.org/W4416615452", "doi": "https://doi.org/10.48550/arxiv.2510.19842", "title": "DAG-Math: Graph-of-Thought Guided Mathematical Reasoning in LLMs", "publication_year": 2025, "publication_date": "2025-10-19", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Zhang, Yuanhe", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I39555362", "display_name": "University of Warwick", "ror": "https://ror.org/01a77tt86", "country_code": "GB", "type": "education", "lineage": ["https://openalex.org/I39555362"]}], "countries": ["GB"], "is_corresponding": false, "raw_author_name": "Zhang, Yuanhe", "raw_affiliation_strings": ["Department of Statistics , University of Warwick , United Kingdom;"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Department of Statistics , University of Warwick , United Kingdom;", "institution_ids": ["https://openalex.org/I39555362"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5041854594", "display_name": "Ilja Kuzborskij", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4210090411", "display_name": "Google DeepMind (United Kingdom)", "ror": "https://ror.org/00971b260", "country_code": "GB", "type": "company", "lineage": ["https://openalex.org/I4210090411", "https://openalex.org/I4210128969"]}], "countries": ["GB"], "is_corresponding": false, "raw_author_name": "Kuzborskij, Ilja", "raw_affiliation_strings": ["Google DeepMind , United Kingdom;"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Google DeepMind , United Kingdom;", "institution_ids": ["https://openalex.org/I4210090411"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5059740024", "display_name": "Jason D. Lee", "orcid": "https://orcid.org/0000-0003-0064-7800", "observed_orcids": ["https://orcid.org/0000-0003-0064-7800"]}, "institutions": [{"id": "https://openalex.org/I39555362", "display_name": "University of Warwick", "ror": "https://ror.org/01a77tt86", "country_code": "GB", "type": "education", "lineage": ["https://openalex.org/I39555362"]}, {"id": "https://openalex.org/I95457486", "display_name": "University of California, Berkeley", "ror": "https://ror.org/01an7q238", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I95457486"]}], "countries": ["GB", "US"], "is_corresponding": false, "raw_author_name": "Lee, Jason D.", "raw_affiliation_strings": ["Department of Electrical Engineering and Computer Sciences , also Department of Statistics , University of California , Berkeley , USA;", "ment of Computer Science , and Centre for Discrete Mathematics and its Applications (DIMAP) , University of Warwick , United Kingdom;"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Department of Electrical Engineering and Computer Sciences , also Department of Statistics , University of California , Berkeley , USA;", "institution_ids": ["https://openalex.org/I95457486"]}, {"raw_affiliation_string": "ment of Computer Science , and Centre for Discrete Mathematics and its Applications (DIMAP) , University of Warwick , United Kingdom;", "institution_ids": ["https://openalex.org/I39555362"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5042142807", "display_name": "Chenlei Leng", "orcid": "https://orcid.org/0000-0001-5703-9617", "observed_orcids": ["https://orcid.org/0000-0001-5703-9617"]}, "institutions": [{"id": "https://openalex.org/I14243506", "display_name": "Hong Kong Polytechnic University", "ror": "https://ror.org/0030zas98", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I14243506"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Leng, Chenlei", "raw_affiliation_strings": ["Department of Applied Mathematics , Hong Kong Polytechnic University , Hong Kong ;"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Department of Applied Mathematics , Hong Kong Polytechnic University , Hong Kong ;", "institution_ids": ["https://openalex.org/I14243506"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5016703530", "display_name": "Fanghui Liu", "orcid": "https://orcid.org/0000-0003-4133-7921", "observed_orcids": ["https://orcid.org/0000-0003-4133-7921"]}, "institutions": [{"id": "https://openalex.org/I4210158445", "display_name": "Institute of Natural Science", "ror": "https://ror.org/004tze884", "country_code": "KP", "type": "education", "lineage": ["https://openalex.org/I4210158445"]}], "countries": ["KP"], "is_corresponding": false, "raw_author_name": "Liu, Fanghui", "raw_affiliation_strings": ["School of Mathematical Sciences , Institute of Natural Sciences and MOE-LSC , Shanghai Jiao Tong University , China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "School of Mathematical Sciences , Institute of Natural Sciences and MOE-LSC , Shanghai Jiao Tong University , China", "institution_ids": ["https://openalex.org/I4210158445"]}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "demonstrate": [4], "strong": [5], "performance": [6], "on": [7, 93], "mathematical": [8, 128], "problems": [9], "when": [10, 144], "prompted": [11], "with": [12], "Chain-of-Thought": [13], "(CoT),": [14], "yet": [15], "it": [16], "remains": [17], "unclear": [18], "whether": [19], "this": [20, 60, 113], "success": [21], "stems": [22], "from": [23], "search,": [24], "rote": [25], "procedures,": [26], "or": [27], "rule-consistent": [28, 153], "reasoning.": [29], "To": [30], "address": [31], "this,": [32, 94], "we": [33, 62, 95], "propose": [34], "modeling": [35], "CoT": [36, 74, 99, 110, 162], "as": [37], "a": [38, 66, 72, 103, 158], "certain": [39], "rule-based": [40], "stochastic": [41], "process": [42], "over": [43], "directed": [44], "acyclic": [45], "graphs": [46], "(DAGs),": [47], "where": [48], "nodes": [49], "represent": [50], "intermediate": [51], "derivation": [52], "states": [53], "and": [54, 101, 152, 163, 176], "edges": [55], "encode": [56], "rule": [57], "applications.": [58], "Within": [59], "framework,": [61], "introduce": [63, 96], "\\textbf{logical": [64], "closeness},": [65], "metric": [67], "that": [68, 105], "quantifies": [69], "how": [70], "well": [71], "model's": [73], "trajectory": [75], "(i.e.,": [76], "the": [77, 83, 97, 117], "LLM's": [78], "final": [79], "output)": [80], "adheres": [81], "to": [82, 108], "DAG": [84], "structure,": [85], "providing": [86], "evaluation": [87, 118], "beyond": [88], "classical": [89], "PASS@$k$": [90, 145], "metrics.": [91], "Building": [92], "\\emph{DAG-MATH}": [98], "format": [100], "construct": [102], "benchmark": [104, 175], "guides": [106], "LLMs": [107, 171], "generate": [109], "trajectories": [111], "in": [112, 137], "format,": [114], "thereby": [115], "enabling": [116], "of": [119], "their": [120], "reasoning": [121, 129, 138, 172], "ability": [122], "under": [123], "our": [124, 131], "framework.": [125], "Across": [126], "standard": [127], "datasets,": [130], "analysis": [132], "uncovers": [133], "statistically": [134], "significant": [135], "differences": [136], "fidelity": [139], "among": [140], "representative": [141], "LLM": [142], "families-even": [143], "is": [146], "comparable-highlighting": [147], "gaps": [148], "between": [149, 160], "final-answer": [150], "accuracy": [151], "derivation.": [154], "Our": [155, 174], "framework": [156], "provides": [157], "balance": [159], "free-form": [161], "formal": [164], "proofs": [165], "systems,": [166], "offering": [167], "actionable": [168], "diagnostics": [169], "for": [170], "evaluation.": [173], "code": [177], "are": [178], "available": [179], "at": [180], "https://github.com/YuanheZ/DAG-MATH.": [181]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5121999979019165, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5121999979019165, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.10499999672174454, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.07850000262260437, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.19842", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.19842", "pdf_url": "https://arxiv.org/pdf/2510.19842", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4408771256", "doi": "https://doi.org/10.2139/ssrn.5107732", "title": "The Mathematics of DeepSeek-R1: Theoretical Foundations and Comparative Analysis", "publication_year": 2025, "publication_date": "2025-01-01", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5007301330", "display_name": "Miquel Noguer I Alonso", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Miquel Noguer I Alonso", "raw_affiliation_strings": ["Artificial Intelligence in Finance Institute", "Artificial Intelligence in Finance Institute, New York, United States"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Artificial Intelligence in Finance Institute", "institution_ids": []}, {"raw_affiliation_string": "Artificial Intelligence in Finance Institute, New York, United States", "institution_ids": []}]}], "abstract_inverted_index": null, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5181000232696533, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5181000232696533, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.11739999800920486, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.10980000346899033, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.2139/ssrn.5107732", "is_oa": true, "landing_page_url": "https://doi.org/10.2139/ssrn.5107732", "pdf_url": null, "source": {"id": "https://openalex.org/S4210172589", "display_name": "SSRN Electronic Journal", "issn_l": "1556-5068", "issn": ["1556-5068"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I1318003438", "host_organization_name": "RELX Group (Netherlands)", "host_organization_lineage": ["https://openalex.org/I1318003438"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "acceptedVersion", "is_accepted": true, "is_published": false, "raw_source_name": null, "raw_type": "posted-content"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4407807310", "doi": "https://doi.org/10.48550/arxiv.2502.14634", "title": "CER: Confidence Enhanced Reasoning in LLMs", "publication_year": 2025, "publication_date": "2025-02-20", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5116361792", "display_name": "Ali Razghandi", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Razghandi, Ali", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5116361793", "display_name": "Seyed Mohammad Hadi Hosseini", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Hosseini, Seyed Mohammad Hadi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5069082023", "display_name": "Mahdieh Soleymani Baghshah", "orcid": "https://orcid.org/0000-0002-1971-6231", "observed_orcids": ["https://orcid.org/0000-0002-1971-6231"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Baghshah, Mahdieh Soleymani", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Ensuring": [0], "the": [1, 39, 65, 84, 104, 115, 156], "reliability": [2, 116], "of": [3, 41, 67, 87, 96, 106, 117, 158, 169], "Large": [4], "Language": [5], "Models": [6], "(LLMs)": [7], "in": [8, 17, 61, 74, 80, 110, 125, 139, 178], "complex": [9], "reasoning": [10, 60, 76, 89], "tasks": [11], "remains": [12], "a": [13, 111], "formidable": [14], "challenge,": [15], "particularly": [16], "scenarios": [18], "that": [19, 57, 113], "demand": [20], "precise": [21], "mathematical": [22, 75, 143], "calculations": [23], "and": [24, 63, 77, 145, 173, 180], "knowledge-intensive": [25], "open-domain": [26, 81, 147, 181], "generation.": [27, 82], "In": [28], "this": [29], "work,": [30], "we": [31, 102], "introduce": [32], "an": [33, 55, 166], "uncertainty-aware": [34], "framework": [35], "designed": [36], "to": [37, 123, 132, 165, 171], "enhance": [38], "accuracy": [40, 167], "LLM": [42], "responses": [43], "by": [44], "systematically": [45], "incorporating": [46], "model": [47], "confidence": [48, 66, 86, 95, 161], "at": [49, 189], "critical": [50, 98], "decision": [51], "points.": [52], "We": [53, 135], "propose": [54], "approach": [56], "encourages": [58], "multi-step": [59], "LLMs": [62], "quantify": [64], "intermediate": [68, 99], "answers": [69], "such": [70], "as": [71], "numerical": [72], "results": [73, 153], "proper": [78], "nouns": [79], "Then,": [83], "overall": [85], "each": [88, 118, 127], "chain": [90, 129], "is": [91, 186], "evaluated": [92], "based": [93], "on": [94], "these": [97], "steps.": [100], "Finally,": [101], "aggregate": [103], "answer": [105], "generated": [107, 119, 128], "response": [108], "paths": [109], "way": [112], "reflects": [114], "content": [120], "(as": [121], "opposed": [122], "self-consistency": [124], "which": [126], "contributes": [130], "equally": [131], "majority": [133], "voting).": [134], "conducted": [136], "extensive": [137], "experiments": [138], "five": [140], "datasets,": [141, 148], "three": [142], "datasets": [144], "two": [146], "using": [149], "four": [150], "LLMs.": [151], "The": [152], "consistently": [154], "validate": [155], "effectiveness": [157], "our": [159], "novel": [160], "aggregation": [162], "method,": [163], "leading": [164], "improvement": [168], "up": [170], "7.4%": [172], "5.8%": [174], "over": [175], "baseline": [176], "approaches": [177], "math": [179], "generation": [182], "tasks,": [183], "respectively.": [184], "Code": [185], "publicly": [187], "available": [188], "https://github.com/": [190], "Aquasar11/CER.": [191]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8438000082969666, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8438000082969666, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.07010000199079514, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11689", "display_name": "Adversarial Robustness in Machine Learning", "score": 0.039500001817941666, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2502.14634", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2502.14634", "pdf_url": "https://arxiv.org/pdf/2502.14634", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414111784", "doi": "https://doi.org/10.1016/j.tsc.2025.102001", "title": "How collaborative problem solving promotes higher-order thinking skills: A systematic review of design features and processes", "publication_year": 2025, "publication_date": "2025-09-10", "type": "review", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5093757464", "display_name": "Hayde\u00e9 Ceballos", "orcid": "https://orcid.org/0000-0001-5141-9651", "observed_orcids": ["https://orcid.org/0000-0001-5141-9651"]}, "institutions": [{"id": "https://openalex.org/I173063890", "display_name": "University of Applied Sciences Utrecht", "ror": "https://ror.org/028z9kw20", "country_code": "NL", "type": "education", "lineage": ["https://openalex.org/I173063890"]}, {"id": "https://openalex.org/I193662353", "display_name": "Utrecht University", "ror": "https://ror.org/04pp8hn57", "country_code": "NL", "type": "education", "lineage": ["https://openalex.org/I193662353"]}], "countries": ["NL"], "is_corresponding": true, "raw_author_name": "Hayde\u00e9 Ceballos", "raw_affiliation_strings": ["Freudenthal Institute, Utrecht University, PO Box 85.170, 3508 AD, Utrecht, the Netherlands", "Hogeschool Utrecht, University of Applied Sciences, PO Box 14007, 3508 SB, Utrecht, the Netherlands"], "raw_orcid": "https://orcid.org/0000-0001-5141-9651", "affiliations": [{"raw_affiliation_string": "Freudenthal Institute, Utrecht University, PO Box 85.170, 3508 AD, Utrecht, the Netherlands", "institution_ids": ["https://openalex.org/I193662353"]}, {"raw_affiliation_string": "Hogeschool Utrecht, University of Applied Sciences, PO Box 14007, 3508 SB, Utrecht, the Netherlands", "institution_ids": ["https://openalex.org/I173063890"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5047039347", "display_name": "Theo van den Bogaart", "orcid": "https://orcid.org/0000-0003-1396-9083", "observed_orcids": ["https://orcid.org/0000-0003-1396-9083"]}, "institutions": [{"id": "https://openalex.org/I173063890", "display_name": "University of Applied Sciences Utrecht", "ror": "https://ror.org/028z9kw20", "country_code": "NL", "type": "education", "lineage": ["https://openalex.org/I173063890"]}], "countries": ["NL"], "is_corresponding": false, "raw_author_name": "Theo van den Bogaart", "raw_affiliation_strings": ["Hogeschool Utrecht, University of Applied Sciences, PO Box 14007, 3508 SB, Utrecht, the Netherlands"], "raw_orcid": "https://orcid.org/0000-0003-1396-9083", "affiliations": [{"raw_affiliation_string": "Hogeschool Utrecht, University of Applied Sciences, PO Box 14007, 3508 SB, Utrecht, the Netherlands", "institution_ids": ["https://openalex.org/I173063890"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5021998527", "display_name": "Stan O. van Ginkel", "orcid": "https://orcid.org/0000-0001-9520-4652", "observed_orcids": ["https://orcid.org/0000-0001-9520-4652"]}, "institutions": [{"id": "https://openalex.org/I173063890", "display_name": "University of Applied Sciences Utrecht", "ror": "https://ror.org/028z9kw20", "country_code": "NL", "type": "education", "lineage": ["https://openalex.org/I173063890"]}], "countries": ["NL"], "is_corresponding": false, "raw_author_name": "Stan van Ginkel", "raw_affiliation_strings": ["Hogeschool Utrecht, University of Applied Sciences, PO Box 14007, 3508 SB, Utrecht, the Netherlands"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Hogeschool Utrecht, University of Applied Sciences, PO Box 14007, 3508 SB, Utrecht, the Netherlands", "institution_ids": ["https://openalex.org/I173063890"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5026915316", "display_name": "Jeroen Spandaw", "orcid": "https://orcid.org/0000-0001-8793-1622", "observed_orcids": ["https://orcid.org/0000-0001-8793-1622"]}, "institutions": [{"id": "https://openalex.org/I98358874", "display_name": "Delft University of Technology", "ror": "https://ror.org/02e2c7k09", "country_code": "NL", "type": "education", "lineage": ["https://openalex.org/I98358874"]}], "countries": ["NL"], "is_corresponding": false, "raw_author_name": "Jeroen Spandaw", "raw_affiliation_strings": ["EWI, Delft University of Technology, PO Box 5031, 2600 GA, Delft, the Netherlands"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "EWI, Delft University of Technology, PO Box 5031, 2600 GA, Delft, the Netherlands", "institution_ids": ["https://openalex.org/I98358874"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5085553542", "display_name": "Paul H. M. Drijvers", "orcid": "https://orcid.org/0000-0002-2724-4967", "observed_orcids": ["https://orcid.org/0000-0002-2724-4967"]}, "institutions": [{"id": "https://openalex.org/I193662353", "display_name": "Utrecht University", "ror": "https://ror.org/04pp8hn57", "country_code": "NL", "type": "education", "lineage": ["https://openalex.org/I193662353"]}], "countries": ["NL"], "is_corresponding": false, "raw_author_name": "Paul Drijvers", "raw_affiliation_strings": ["Freudenthal Institute, Utrecht University, PO Box 85.170, 3508 AD, Utrecht, the Netherlands"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Freudenthal Institute, Utrecht University, PO Box 85.170, 3508 AD, Utrecht, the Netherlands", "institution_ids": ["https://openalex.org/I193662353"]}]}], "abstract_inverted_index": {"\u2022": [0, 11, 22, 32, 42], "Novel": [1], "language-based": [2], "machine": [3, 131], "learning": [4, 89, 132, 170], "technique": [5], "used": [6, 129], "in": [7, 14, 57, 66, 163, 280], "this": [8, 268], "literature": [9], "review.": [10], "Using": [12], "technology": [13], "geometry": [15], "tasks": [16, 181, 260], "enhances": [17], "students'": [18], "mathematical": [19], "reasoning": [20], "skills.": [21, 51, 302], "Open-ended": [23], "problems": [24], "boost": [25], "reflective": [26, 186], "discourse,": [27], "teamwork,": [28], "and": [29, 79, 88, 93, 114, 122, 138, 171, 182, 188, 202, 212, 239, 254, 275, 296], "cognitive": [30, 161, 190, 301], "processes.": [31], "Students'": [33], "communication": [34, 288], "quality": [35, 44], "is": [36, 289], "crucial": [37], "for": [38, 63, 273, 291], "collaborative": [39, 294], "problem-solving": [40], "success.": [41], "Communication": [43], "impacts": [45], "the": [46, 140, 154, 230, 297], "development": [47, 298], "of": [48, 148, 169, 232, 299], "higher-order": [49], "thinking": [50, 78], "Collaborative": [52], "Problem": [53], "Solving": [54], "(CPS)": [55], "work": [56], "mathematics": [58, 281], "education": [59], "are": [60], "widely": [61], "recognized": [62], "engaging": [64], "students": [65], "cognitively": [67], "demanding": [68], "activities": [69], "that": [70, 195, 284], "foster": [71], "Higher-Order": [72], "Thinking": [73], "Skills": [74], "(HOTS),": [75], "like": [76, 210], "critical": [77], "reasoning.": [80], "However,": [81], "connections": [82], "between": [83, 112], "design": [84, 175, 235], "features,": [85, 176], "CPS": [86, 123, 164, 217, 237, 259, 278], "processes,": [87], "outcomes": [90], "remain": [91], "complex": [92], "not": [94], "fully": [95, 219], "understood.": [96], "To": [97], "address": [98, 243], "this,": [99], "we": [100], "applied": [101], "a": [102, 130, 145], "conjecture-based": [103, 155], "framework": [104], "to": [105, 126, 134, 223, 242, 249, 263], "systematically": [106], "review": [107, 269], "45": [108], "empirical": [109], "studies": [110, 137], "published": [111], "2010": [113], "2022,": [115], "focusing": [116], "on": [117], "how": [118, 160], "specific": [119], "task": [120, 234], "designs": [121], "processes": [124, 162, 238, 295], "contribute": [125], "HOTS.": [127, 225, 265], "We": [128, 192], "tool": [133], "prioritize": [135], "relevant": [136], "streamline": [139], "selection": [141], "process,": [142], "ending": [143], "after": [144], "threshold": [146], "number": [147], "consecutive": [149], "irrelevant": [150], "articles.": [151], "Guided": [152], "by": [153], "framework,": [156], "our": [157], "analysis": [158], "highlighted": [159], "function": [165], "as": [166, 178], "essential": [167], "mechanisms": [168], "measurable": [172], "outcomes.": [173], "Specifically,": [174], "such": [177], "technology-supported": [179], "exploratory": [180], "open-ended": [183], "problems,": [184], "encourage": [185], "discourse": [187], "deeper": [189], "engagement.": [191], "also": [193], "found": [194], "structured": [196, 287], "group": [197], "procedures,": [198], "including": [199], "clear": [200, 252], "roles": [201], "guided": [203], "interaction": [204], "protocols,": [205], "improve": [206], "collaboration.": [207], "Nonetheless,": [208], "challenges": [209], "miscommunication": [211], "uneven": [213], "participation": [214], "can": [215, 261], "limit": [216], "from": [218], "realizing": [220], "its": [221], "potential": [222], "cultivate": [224, 264], "Overall,": [226], "these": [227], "findings": [228], "underscore": [229], "importance": [231], "aligning": [233], "with": [236], "using": [240], "strategies": [241], "collaboration": [244], "barriers,": [245], "particularly": [246], "those": [247], "related": [248], "communication.": [250], "Without": [251], "protocols": [253], "consistent": [255], "dialogue,": [256], "even": [257], "well-designed": [258], "fail": [262], "In": [266], "conclusion,": [267], "offers": [270], "practical": [271], "insights": [272], "educators": [274], "researchers": [276], "implementing": [277], "effectively": [279], "education,": [282], "highlighting": [283], "fostering": [285], "open,": [286], "vital": [290], "optimizing": [292], "both": [293], "advanced": [300]}, "primary_topic": {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.9761000275611877, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.9761000275611877, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10636", "display_name": "Innovative Teaching and Learning Methods", "score": 0.01679999940097332, "subfield": {"id": "https://openalex.org/subfields/3204", "display_name": "Developmental and Educational Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T10130", "display_name": "Mathematics Education and Teaching Techniques", "score": 0.0026000000070780516, "subfield": {"id": "https://openalex.org/subfields/3304", "display_name": "Education"}, "field": {"id": "https://openalex.org/fields/33", "display_name": "Social Sciences"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}], "primary_location": {"id": "doi:10.1016/j.tsc.2025.102001", "is_oa": true, "landing_page_url": "https://doi.org/10.1016/j.tsc.2025.102001", "pdf_url": null, "source": {"id": "https://openalex.org/S42886211", "display_name": "Thinking Skills and Creativity", "issn_l": "1871-1871", "issn": ["1871-1871", "1878-0423"], "is_oa": false, "is_in_doaj": false, "is_core": true, "listed_in": ["cwts-core", "erih-plus", "jufo-1", "norway-1"], "host_organization": "https://openalex.org/P4310320990", "host_organization_name": "Elsevier BV", "host_organization_lineage": ["https://openalex.org/P4310320990"], "host_organization_lineage_names": ["Elsevier BV"], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Thinking Skills and Creativity", "raw_type": "journal-article"}, "cited_by_count": 19, "is_retracted": false}, {"id": "https://openalex.org/W4416260813", "doi": "https://doi.org/10.48550/arxiv.2506.22005", "title": "LeanConjecturer: Automatic Generation of Mathematical Conjectures for Theorem Proving", "publication_year": 2025, "publication_date": "2025-06-27", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5058374581", "display_name": "Naoto Onda", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Onda, Naoto", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5083407612", "display_name": "Kazumi Kasaura", "orcid": "https://orcid.org/0000-0002-3219-9961", "observed_orcids": ["https://orcid.org/0000-0002-3219-9961"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kasaura, Kazumi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5140269507", "display_name": "Yuta Oriike", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Oriike, Yuta", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Taniguchi, Masaya", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Taniguchi, Masaya", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5056836344", "display_name": "Akiyoshi Sannai", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Sannai, Akiyoshi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5036537988", "display_name": "Sho Sonoda", "orcid": "https://orcid.org/0000-0001-7242-4740", "observed_orcids": ["https://orcid.org/0000-0001-7242-4740"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Sonoda, Sho", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"We": [0, 70], "introduce": [1], "LeanConjecturer,": [2], "a": [3, 111], "pipeline": [4], "for": [5, 78, 114, 118, 142], "automatically": [6], "generating": [7], "university-level": [8], "mathematical": [9, 143], "conjectures": [10, 48, 77, 93, 104], "in": [11, 36, 129], "Lean": [12], "4": [13], "using": [14], "Large": [15], "Language": [16], "Models": [17], "(LLMs).": [18], "Our": [19, 99, 122], "hybrid": [20], "approach": [21, 100], "combines": [22], "rule-based": [23], "context": [24], "extraction": [25], "with": [26, 54], "LLM-based": [27], "theorem": [28, 38, 96, 119], "statement": [29], "generation,": [30], "addressing": [31], "the": [32, 72], "data": [33, 117], "scarcity": [34], "challenge": [35], "formal": [37], "proving.": [39], "Through": [40], "iterative": [41], "generation": [42], "and": [43, 60, 136], "evaluation,": [44], "LeanConjecturer": [45], "produced": [46], "12,289": [47], "from": [49], "40": [50], "Mathlib": [51], "seed": [52, 106], "files,": [53], "3,776": [55], "identified": [56], "as": [57], "syntactically": [58], "valid": [59], "non-trivial,": [61], "that": [62, 88], "is,": [63], "cannot": [64], "be": [65], "proven": [66], "by": [67], "\\texttt{aesop}": [68], "tactic.": [69], "demonstrate": [71], "utility": [73], "of": [74, 133, 148], "these": [75], "generated": [76], "reinforcement": [79], "learning": [80], "through": [81], "Group": [82], "Relative": [83], "Policy": [84], "Optimization": [85], "(GRPO),": [86], "showing": [87], "targeted": [89], "training": [90, 116], "on": [91, 108], "domain-specific": [92], "can": [94], "enhance": [95], "proving": [97, 120], "capabilities.": [98], "generates": [101], "103.25": [102], "novel": [103], "per": [105], "file": [107], "average,": [109], "providing": [110], "scalable": [112], "solution": [113], "creating": [115], "systems.": [121], "system": [123], "successfully": [124], "verified": [125], "several": [126], "non-trivial": [127], "theorems": [128], "topology,": [130], "including": [131], "properties": [132], "semi-open,": [134], "alpha-open,": [135], "pre-open": [137], "sets,": [138], "demonstrating": [139], "its": [140], "potential": [141], "discovery": [144], "beyond": [145], "simple": [146], "variations": [147], "existing": [149], "results.": [150]}, "primary_topic": {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.42239999771118164, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.42239999771118164, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.4140999913215637, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.03759999945759773, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2506.22005", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2506.22005", "pdf_url": "https://arxiv.org/pdf/2506.22005", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417250113", "doi": "https://doi.org/10.48550/arxiv.2504.07491", "title": "Kimi-VL Technical Report", "publication_year": 2025, "publication_date": "2025-04-10", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5116011620", "display_name": "Kimi Team", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kimi Team", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5088792207", "display_name": "Angang Du", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Du, Angang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Yin, Bohong", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yin, Bohong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5028438544", "display_name": "Bowei Xing", "orcid": "https://orcid.org/0009-0001-3254-3902", "observed_orcids": ["https://orcid.org/0009-0001-3254-3902"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xing, Bowei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Qu, Bowen", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Qu, Bowen", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100412535", "display_name": "Bowen Wang", "orcid": "https://orcid.org/0000-0001-5794-2144", "observed_orcids": ["https://orcid.org/0000-0001-5794-2144"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Bowen", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101559656", "display_name": "Cheng Chen", "orcid": "https://orcid.org/0000-0002-0586-0646", "observed_orcids": ["https://orcid.org/0000-0002-0586-0646"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Cheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5103265154", "display_name": "Chenlin Zhang", "orcid": "https://orcid.org/0000-0001-5116-4046", "observed_orcids": ["https://orcid.org/0000-0001-5116-4046"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Chenlin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5044355459", "display_name": "Chenzhuang Du", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Du, Chenzhuang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Wei, Chu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wei, Chu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5111099880", "display_name": "Congcong Wang", "orcid": "https://orcid.org/0009-0006-5330-4691", "observed_orcids": ["https://orcid.org/0009-0006-5330-4691"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Congcong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5046162247", "display_name": "Dehao Zhang", "orcid": "https://orcid.org/0009-0007-4493-5595", "observed_orcids": ["https://orcid.org/0009-0007-4493-5595"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Dehao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Du, Dikang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Du, Dikang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100689748", "display_name": "Dongliang Wang", "orcid": "https://orcid.org/0000-0001-7052-4182", "observed_orcids": ["https://orcid.org/0000-0001-7052-4182"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Dongliang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5073001520", "display_name": "Enming Yuan", "orcid": "https://orcid.org/0000-0003-0755-9109", "observed_orcids": ["https://orcid.org/0000-0003-0755-9109"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yuan, Enming", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5049660040", "display_name": "Enzhe Lu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Lu, Enzhe", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5020047491", "display_name": "Fang Li", "orcid": "https://orcid.org/0000-0002-6908-7920", "observed_orcids": ["https://orcid.org/0000-0002-6908-7920"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Fang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5065232789", "display_name": "Flood Sung", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Sung, Flood", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Wei, Guangda", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wei, Guangda", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5015579277", "display_name": "Guokun Lai", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Lai, Guokun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5063667378", "display_name": "Zhu Han", "orcid": "https://orcid.org/0000-0002-6606-5822", "observed_orcids": ["https://orcid.org/0000-0002-6606-5822"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhu, Han", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101528531", "display_name": "Hao Ding", "orcid": "https://orcid.org/0000-0003-2688-3852", "observed_orcids": ["https://orcid.org/0000-0003-2688-3852"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ding, Hao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5046973045", "display_name": "Hao Hu", "orcid": "https://orcid.org/0000-0002-3567-609X", "observed_orcids": ["https://orcid.org/0000-0002-3567-609X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Hu, Hao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101457978", "display_name": "Hao Yang", "orcid": "https://orcid.org/0009-0005-8623-8531", "observed_orcids": ["https://orcid.org/0009-0005-8623-8531"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yang, Hao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5082017754", "display_name": "Hao Zhang", "orcid": "https://orcid.org/0000-0002-0980-8888", "observed_orcids": ["https://orcid.org/0000-0002-0980-8888"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Hao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5039135087", "display_name": "Haoning Wu", "orcid": "https://orcid.org/0009-0001-8717-338X", "observed_orcids": ["https://orcid.org/0009-0001-8717-338X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wu, Haoning", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5071838616", "display_name": "Haotian Yao", "orcid": "https://orcid.org/0000-0003-4674-1552", "observed_orcids": ["https://orcid.org/0000-0003-4674-1552"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yao, Haotian", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5083840550", "display_name": "H. Q. Lu", "orcid": "https://orcid.org/0000-0001-7759-9443", "observed_orcids": ["https://orcid.org/0000-0001-7759-9443"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Lu, Haoyu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100453945", "display_name": "Heng Wang", "orcid": "https://orcid.org/0000-0001-9338-6624", "observed_orcids": ["https://orcid.org/0000-0001-9338-6624"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Heng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5110370692", "display_name": "Hongcheng Gao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Gao, Hongcheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5111649512", "display_name": "Huabin Zheng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zheng, Huabin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100326742", "display_name": "Jiaming Li", "orcid": "https://orcid.org/0000-0002-0109-0867", "observed_orcids": ["https://orcid.org/0000-0002-0109-0867", "https://orcid.org/0000-0002-6424-1554"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Jiaming", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5078222384", "display_name": "Jianlin Su", "orcid": "https://orcid.org/0000-0002-1000-8458", "observed_orcids": ["https://orcid.org/0000-0002-1000-8458"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Su, Jianlin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5044720245", "display_name": "Jianzhou Wang", "orcid": "https://orcid.org/0000-0001-9078-7617", "observed_orcids": ["https://orcid.org/0000-0001-9078-7617", "https://orcid.org/0000-0001-6070-9888"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Jianzhou", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5062194013", "display_name": "Jiaqi Deng", "orcid": "https://orcid.org/0009-0001-1741-9882", "observed_orcids": ["https://orcid.org/0009-0001-1741-9882"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Deng, Jiaqi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5083333630", "display_name": "Jiezhong Qiu", "orcid": "https://orcid.org/0000-0001-9514-0708", "observed_orcids": ["https://orcid.org/0000-0001-9514-0708"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Qiu, Jiezhong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5035415865", "display_name": "Jin Xie", "orcid": "https://orcid.org/0000-0001-8523-6052", "observed_orcids": ["https://orcid.org/0000-0001-8523-6052"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xie, Jin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100775334", "display_name": "Jinhong Wang", "orcid": "https://orcid.org/0000-0002-0809-9856", "observed_orcids": ["https://orcid.org/0000-0002-0809-9856"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Jinhong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Liu, Jingyuan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Jingyuan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5103205656", "display_name": "Junjie Yan", "orcid": "https://orcid.org/0009-0001-7918-3801", "observed_orcids": ["https://orcid.org/0009-0001-7918-3801"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yan, Junjie", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Ouyang, Kun", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ouyang, Kun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100604810", "display_name": "Liang Chen", "orcid": "https://orcid.org/0000-0001-5006-4094", "observed_orcids": ["https://orcid.org/0000-0001-5006-4094"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Liang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Sui, Lin", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Sui, Lin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5049419584", "display_name": "Longhui Yu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yu, Longhui", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5072164184", "display_name": "Mengfan Dong", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Dong, Mengfan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Dong, Mengnan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Dong, Mengnan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101486827", "display_name": "Nuo Xu", "orcid": "https://orcid.org/0000-0003-2586-5804", "observed_orcids": ["https://orcid.org/0000-0003-2586-5804"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Nuo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5110618868", "display_name": "Pengyu Cheng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Cheng, Pengyu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Gu, Qizheng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Gu, Qizheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5113209710", "display_name": "Runjie Zhou", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhou, Runjie", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102884727", "display_name": "Shaowei Liu", "orcid": "https://orcid.org/0009-0004-6105-7999", "observed_orcids": ["https://orcid.org/0009-0004-6105-7999"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Shaowei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Cao, Sihan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Cao, Sihan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100596110", "display_name": "Tao Yu", "orcid": "https://orcid.org/0009-0001-8083-2418", "observed_orcids": ["https://orcid.org/0009-0001-8083-2418"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yu, Tao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5091391590", "display_name": "Tianhui Song", "orcid": "https://orcid.org/0009-0007-4256-6407", "observed_orcids": ["https://orcid.org/0009-0007-4256-6407"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Song, Tianhui", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Bai, Tongtong", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Bai, Tongtong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5019370942", "display_name": "Wei Song", "orcid": "https://orcid.org/0000-0002-4780-0350", "observed_orcids": ["https://orcid.org/0000-0002-4780-0350"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Song, Wei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "He, Weiran", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "He, Weiran", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5013900180", "display_name": "Weixiao Huang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Huang, Weixiao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5109953055", "display_name": "Weixin Xu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Weixin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5071207563", "display_name": "Xiaokun Yuan", "orcid": "https://orcid.org/0000-0003-1515-8946", "observed_orcids": ["https://orcid.org/0000-0003-1515-8946"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yuan, Xiaokun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5065739587", "display_name": "Xingcheng Yao", "orcid": "https://orcid.org/0000-0001-5065-3016", "observed_orcids": ["https://orcid.org/0000-0001-5065-3016"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yao, Xingcheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5086375722", "display_name": "Xingzhe Wu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wu, Xingzhe", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5038731476", "display_name": "Xinhao Li", "orcid": "https://orcid.org/0000-0002-6819-5460", "observed_orcids": ["https://orcid.org/0000-0002-6819-5460"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Xinhao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5028097242", "display_name": "Xinxing Zu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zu, Xinxing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5074959378", "display_name": "Xinyu Zhou", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhou, Xinyu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5054558475", "display_name": "Xinyuan Wang", "orcid": "https://orcid.org/0000-0002-8572-0407", "observed_orcids": ["https://orcid.org/0000-0002-8572-0407"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Xinyuan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5076124480", "display_name": "Yann Philippe Charles", "orcid": "https://orcid.org/0000-0002-0606-211X", "observed_orcids": ["https://orcid.org/0000-0002-0606-211X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Charles, Y.", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5110992737", "display_name": "Yan Zhong", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhong, Yan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100460352", "display_name": "Li Yang", "orcid": "https://orcid.org/0000-0003-4791-1349", "observed_orcids": ["https://orcid.org/0000-0003-4791-1349"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Yang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101075955", "display_name": "Yangyang Hu", "orcid": "https://orcid.org/0009-0007-4457-5630", "observed_orcids": ["https://orcid.org/0009-0007-4457-5630"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Hu, Yangyang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100695892", "display_name": "Yanru Chen", "orcid": "https://orcid.org/0000-0002-6531-7995", "observed_orcids": ["https://orcid.org/0000-0002-6531-7995", "https://orcid.org/0009-0003-4135-0488", "https://orcid.org/0009-0004-9330-1028", "https://orcid.org/0009-0007-1647-8999"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Yanru", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5104243967", "display_name": "Yejie Wang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Yejie", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Liu, Yibo", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Yibo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5045257498", "display_name": "Yibo Miao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Miao, Yibo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Qin, Yidao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Qin, Yidao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100377728", "display_name": "Yimin Chen", "orcid": "https://orcid.org/0000-0003-1562-1443", "observed_orcids": ["https://orcid.org/0000-0003-1562-1443", "https://orcid.org/0009-0003-8204-7143"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Yimin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5143926194", "display_name": "Yiping Bao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Bao, Yiping", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100662108", "display_name": "Yiqin Wang", "orcid": "https://orcid.org/0000-0002-2783-8861", "observed_orcids": ["https://orcid.org/0000-0002-2783-8861", "https://orcid.org/0000-0001-8480-2040"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Yiqin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100931742", "display_name": "Yongsheng Kang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kang, Yongsheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5044541656", "display_name": "Yuanxin Liu", "orcid": "https://orcid.org/0000-0002-0673-4887", "observed_orcids": ["https://orcid.org/0000-0002-0673-4887"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Yuanxin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101315785", "display_name": "Yuhao Dong", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Dong, Yuhao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5012998749", "display_name": "Yulun Du", "orcid": "https://orcid.org/0009-0002-6171-6750", "observed_orcids": ["https://orcid.org/0009-0002-6171-6750"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Du, Yulun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Wu, Yuxin", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wu, Yuxin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100669046", "display_name": "Yuzhi Wang", "orcid": "https://orcid.org/0000-0002-9339-2386", "observed_orcids": ["https://orcid.org/0000-0002-9339-2386"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Yuzhi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Yan, Yuzi", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yan, Yuzi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5113154912", "display_name": "Zaida Zhou", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhou, Zaida", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5008593161", "display_name": "Zhaowei Li", "orcid": "https://orcid.org/0000-0002-5176-4655", "observed_orcids": ["https://orcid.org/0000-0002-5176-4655"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Zhaowei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5012056827", "display_name": "Zhejun Jiang", "orcid": "https://orcid.org/0009-0000-0891-7380", "observed_orcids": ["https://orcid.org/0009-0000-0891-7380"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Jiang, Zhejun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100459108", "display_name": "Zheng Zhang", "orcid": "https://orcid.org/0000-0002-2338-6108", "observed_orcids": ["https://orcid.org/0000-0002-2338-6108"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Zheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5050303775", "display_name": "Zhilin Yang", "orcid": "https://orcid.org/0000-0003-0944-4903", "observed_orcids": ["https://orcid.org/0000-0003-0944-4903", "https://orcid.org/0000-0002-8598-922X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yang, Zhilin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5103191012", "display_name": "Zhiqi Huang", "orcid": "https://orcid.org/0000-0002-8317-0966", "observed_orcids": ["https://orcid.org/0000-0002-8317-0966"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Huang, Zhiqi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Huang, Zihao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Huang, Zihao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102012495", "display_name": "Zijia Zhao", "orcid": "https://orcid.org/0009-0005-8242-6092", "observed_orcids": ["https://orcid.org/0009-0005-8242-6092"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhao, Zijia", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5049260052", "display_name": "Ziwei Chen", "orcid": "https://orcid.org/0000-0001-7679-9582", "observed_orcids": ["https://orcid.org/0000-0001-7679-9582"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Ziwei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": null, "display_name": "Lin, Zongyu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Lin, Zongyu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"We": [0], "present": [1], "Kimi-VL,": [2, 168], "an": [3, 171], "efficient": [4, 87], "open-source": [5], "Mixture-of-Experts": [6], "(MoE)": [7], "vision-language": [8], "model": [9, 190], "(VLM)": [10], "that": [11], "offers": [12], "advanced": [13, 172], "multimodal": [14], "reasoning,": [15, 75], "long-context": [16], "understanding,": [17], "and": [18, 70, 76, 93, 109, 131, 145, 154, 184, 217], "strong": [19, 36, 192], "agent": [20, 49], "capabilities": [21, 60, 195], "-": [22], "all": [23], "while": [24, 95, 158, 211], "activating": [25], "only": [26], "2.8B": [27], "parameters": [28], "in": [29, 47, 98, 105], "its": [30], "language": [31, 65], "decoder": [32], "(Kimi-VL-A3B).": [33], "Kimi-VL": [34, 45, 102, 118], "demonstrates": [35], "performance": [37], "across": [38, 61], "challenging": [39, 63], "domains:": [40], "as": [41, 90], "a": [42, 113], "general-purpose": [43], "VLM,": [44], "excels": [46], "multi-turn": [48], "tasks": [50], "(e.g.,": [51], "OSWorld),": [52], "matching": [53], "flagship": [54], "models.": [55], "Furthermore,": [56], "it": [57, 82, 142], "exhibits": [58, 191], "remarkable": [59], "diverse": [62, 121], "vision": [64, 137], "tasks,": [66], "including": [67], "college-level": [68], "image": [69], "video": [71], "comprehension,": [72], "OCR,": [73], "mathematical": [74], "multi-image": [77], "understanding.": [78], "In": [79], "comparative": [80], "evaluations,": [81], "effectively": [83], "competes": [84], "with": [85], "cutting-edge": [86], "VLMs": [88], "such": [89], "GPT-4o-mini,": [91], "Qwen2.5-VL-7B,": [92], "Gemma-3-12B-IT,": [94], "surpassing": [96], "GPT-4o": [97], "several": [99], "key": [100], "domains.": [101], "also": [103], "advances": [104], "processing": [106], "long": [107, 122, 178], "contexts": [108], "perceiving": [110], "clearly.": [111], "With": [112], "128K": [114], "extended": [115], "context": [116], "window,": [117], "can": [119], "process": [120], "inputs,": [123, 149], "achieving": [124, 150], "impressive": [125], "scores": [126], "of": [127], "64.5": [128], "on": [129, 133, 152, 156, 197, 200, 203, 206, 209], "LongVideoBench": [130], "35.1": [132], "MMLongBench-Doc.": [134], "Its": [135], "native-resolution": [136], "encoder,": [138], "MoonViT,": [139], "further": [140], "allows": [141], "to": [143], "see": [144], "understand": [146], "ultra-high-resolution": [147], "visual": [148], "83.2": [151], "InfoVQA": [153], "34.5": [155], "ScreenSpot-Pro,": [157], "maintaining": [159], "lower": [160], "computational": [161], "cost": [162], "for": [163], "common": [164], "tasks.": [165], "Building": [166], "upon": [167], "we": [169], "introduce": [170], "long-thinking": [173], "variant:": [174], "Kimi-VL-Thinking-2506.": [175], "Developed": [176], "through": [177], "chain-of-thought": [179], "(CoT)": [180], "supervised": [181], "fine-tuning": [182], "(SFT)": [183], "reinforcement": [185], "learning": [186], "(RL),": [187], "the": [188], "latest": [189], "long-horizon": [193], "reasoning": [194], "(64.0": [196], "MMMU,": [198], "46.3": [199], "MMMU-Pro,": [201], "56.9": [202], "MathVision,": [204], "80.1": [205], "MathVista,": [207], "65.2": [208], "VideoMMMU)": [210], "obtaining": [212], "robust": [213], "general": [214], "abilities.": [215], "Code": [216], "models": [218], "are": [219], "publicly": [220], "accessible": [221], "at": [222], "https://github.com/MoonshotAI/Kimi-VL.": [223]}, "primary_topic": {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 1.0, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 1.0, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10036", "display_name": "Advanced Neural Network Applications", "score": 0.0, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.0, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2504.07491", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2504.07491", "pdf_url": "https://arxiv.org/pdf/2504.07491", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by-nc-sa", "license_id": "https://openalex.org/licenses/cc-by-nc-sa", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 3, "is_retracted": false}, {"id": "https://openalex.org/W4415150866", "doi": "https://doi.org/10.48550/arxiv.2503.04104", "title": "LLMs Can Generate a Better Answer by Aggregating Their Own Responses", "publication_year": 2025, "publication_date": "2025-03-06", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Li, Zichong", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Zichong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5055356629", "display_name": "Xinyu Feng", "orcid": "https://orcid.org/0000-0002-4374-374X", "observed_orcids": ["https://orcid.org/0000-0002-4374-374X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Feng, Xinyu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Cai, Yuheng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Cai, Yuheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Zhang, Zixuan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Zixuan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5057150343", "display_name": "T Liu", "orcid": "https://orcid.org/0009-0001-1037-1804", "observed_orcids": ["https://orcid.org/0009-0001-1037-1804", "https://orcid.org/0000-0003-1635-2986"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Tianyi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101926199", "display_name": "Chen Liang", "orcid": "https://orcid.org/0000-0003-1781-859X", "observed_orcids": ["https://orcid.org/0000-0003-1781-859X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liang, Chen", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5051745436", "display_name": "Weizhu Chen", "orcid": "https://orcid.org/0000-0003-4196-7229", "observed_orcids": ["https://orcid.org/0000-0003-4196-7229"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Weizhu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5055254609", "display_name": "Haoyu Wang", "orcid": "https://orcid.org/0009-0002-2266-6498", "observed_orcids": ["https://orcid.org/0009-0002-2266-6498"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Haoyu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": null, "display_name": "Zhao, Tuo", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhao, Tuo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "have": [4, 28, 35], "shown": [5, 36], "remarkable": [6], "capabilities": [7], "across": [8, 192], "tasks,": [9, 194], "yet": [10], "they": [11], "often": [12], "require": [13, 119], "additional": [14], "prompting": [15, 83], "techniques": [16], "when": [17, 41], "facing": [18], "complex": [19], "problems.": [20], "While": [21, 148], "approaches": [22], "like": [23], "self-correction": [24], "and": [25, 176, 200, 208], "response": [26, 127, 140, 158, 190], "selection": [27, 51], "emerged": [29], "as": [30, 205], "popular": [31], "solutions,": [32], "recent": [33], "studies": [34], "these": [37], "methods": [38], "perform": [39], "poorly": [40], "relying": [42], "on": [43, 142], "the": [44, 59, 91, 102, 120, 132, 143, 153], "LLM": [45, 63, 121], "itself": [46], "to": [47, 107, 122, 136, 165, 180], "provide": [48], "feedback": [49], "or": [50, 125], "criteria.": [52], "We": [53], "argue": [54], "this": [55, 74], "limitation": [56], "stems": [57], "from": [58, 101], "fact": [60], "that": [61, 85, 186], "common": [62], "post-training": [64], "procedures": [65], "lack": [66], "explicit": [67], "supervision": [68], "for": [69, 157], "discriminative": [70, 93], "judgment": [71], "tasks.": [72, 182], "In": [73, 169], "paper,": [75], "we": [76], "propose": [77], "Generative": [78], "Self-Aggregation": [79], "(GSA),": [80], "a": [81, 138], "novel": [82], "method": [84, 116], "improves": [86, 189], "answer": [87], "quality": [88, 191], "without": [89], "requiring": [90], "model's": [92, 133], "capabilities.": [94], "GSA": [95, 149, 187], "first": [96], "samples": [97], "multiple": [98, 146], "diverse": [99], "responses": [100], "LLM,": [103], "then": [104], "aggregates": [105], "them": [106], "obtain": [108], "an": [109], "improved": [110], "solution.": [111], "Unlike": [112], "previous": [113], "approaches,": [114], "our": [115, 171], "does": [117], "not": [118], "correct": [123], "errors": [124], "compare": [126], "quality;": [128], "instead,": [129], "it": [130], "leverages": [131], "generative": [134], "abilities": [135], "synthesize": [137], "new": [139], "based": [141], "context": [144], "of": [145], "samples.": [147], "shares": [150], "similarities": [151], "with": [152], "self-consistency": [154], "(SC)": [155], "approach": [156, 172], "aggregation,": [159], "SC": [160], "requires": [161], "specific": [162], "verifiable": [163], "tokens": [164], "enable": [166], "majority": [167], "voting.": [168], "contrast,": [170], "is": [173], "more": [174], "general": [175], "can": [177], "be": [178], "applied": [179], "open-ended": [181, 201], "Empirical": [183], "evaluation": [184], "demonstrates": [185], "effectively": [188], "various": [193], "including": [195], "mathematical": [196], "reasoning,": [197], "knowledge-based": [198], "problems,": [199], "generation": [202], "tasks": [203], "such": [204], "code": [206], "synthesis": [207], "conversational": [209], "responses.": [210]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.9375, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.9375, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.04569999873638153, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.004900000058114529, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2503.04104", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2503.04104", "pdf_url": "https://arxiv.org/pdf/2503.04104", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415442622", "doi": "https://doi.org/10.1002/advs.202517294", "title": "A Perspective on Interactive Theorem Provers in Physics", "publication_year": 2025, "publication_date": "2025-10-22", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5030966353", "display_name": "Joseph Tooby\u2010Smith", "orcid": "https://orcid.org/0000-0003-2831-598X", "observed_orcids": ["https://orcid.org/0000-0003-2831-598X"]}, "institutions": [{"id": "https://openalex.org/I51601045", "display_name": "University of Bath", "ror": "https://ror.org/002h8g185", "country_code": "GB", "type": "education", "lineage": ["https://openalex.org/I51601045"]}], "countries": ["GB"], "is_corresponding": true, "raw_author_name": "Joseph Tooby\u2010Smith", "raw_affiliation_strings": ["Department of Computer Science University of Bath  Bath BA2 7AU UK", "Department of Computer Science, University of Bath, Bath, BA2 7AU UK"], "raw_orcid": "https://orcid.org/0000-0003-2831-598X", "affiliations": [{"raw_affiliation_string": "Department of Computer Science University of Bath  Bath BA2 7AU UK", "institution_ids": ["https://openalex.org/I51601045"]}, {"raw_affiliation_string": "Department of Computer Science, University of Bath, Bath, BA2 7AU UK", "institution_ids": ["https://openalex.org/I51601045"]}]}], "abstract_inverted_index": {"Interactive": [0], "theorem": [1], "provers": [2], "(ITPs)": [3], "are": [4, 43], "a": [5], "class": [6], "of": [7, 15, 31, 41, 53], "computer": [8], "programming": [9], "language": [10], "that": [11], "guarantee": [12], "the": [13, 29, 38, 51], "correctness": [14], "mathematical": [16], "definitions": [17], "and": [18, 26, 49], "theorems": [19], "written": [20], "in": [21], "to": [22], "them.": [23], "Within": [24], "mathematics": [25], "artificial": [27], "intelligence": [28], "usage": [30, 40], "ITPs": [32, 42], "is": [33], "growing.": [34], "In": [35], "this": [36], "perspective,": [37], "best": [39], "discussed": [44], "within": [45], "another": [46], "area,": [47], "physics,": [48], "motivate": [50], "existence": [52], "an": [54], "open-source": [55], "community": [56], "run": [57], "project": [58], "for": [59], "formalized": [60], "physics": [61], "results,": [62], "PhysLean.": [63]}, "primary_topic": {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.9955999851226807, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.9955999851226807, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.0015999999595806003, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.0010999999940395355, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.1002/advs.202517294", "is_oa": true, "landing_page_url": "https://doi.org/10.1002/advs.202517294", "pdf_url": "https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/advs.202517294", "source": {"id": "https://openalex.org/S2737737698", "display_name": "Advanced Science", "issn_l": "2198-3844", "issn": ["2198-3844"], "is_oa": true, "is_in_doaj": true, "is_core": true, "listed_in": ["cwts-core", "doaj", "jufo-1", "ki-jl-2", "medline", "norway-1"], "host_organization": "https://openalex.org/P4310320595", "host_organization_name": "Wiley", "host_organization_lineage": ["https://openalex.org/P4310320595"], "host_organization_lineage_names": ["Wiley"], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Advanced Science", "raw_type": "journal-article"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4410768311", "doi": "https://doi.org/10.20944/preprints202505.2064.v1", "title": "A Comprehensive Review of Qwen and DeepSeek LLMs: Architecture, Performance and Applications", "publication_year": 2025, "publication_date": "2025-05-27", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5108368607", "display_name": "Satyadhar Joshi", "orcid": "https://orcid.org/0009-0002-6011-5080", "observed_orcids": ["https://orcid.org/0009-0002-6011-5080"]}, "institutions": [{"id": "https://openalex.org/I13955877", "display_name": "Bar-Ilan University", "ror": "https://ror.org/03kgsv495", "country_code": "IL", "type": "education", "lineage": ["https://openalex.org/I13955877"]}], "countries": ["IL"], "is_corresponding": true, "raw_author_name": "Satyadhar Joshi", "raw_affiliation_strings": ["Independent , Alumnus , International MBA , Bar-Ilan University , Israel;"], "raw_orcid": "https://orcid.org/0009-0002-6011-5080", "affiliations": [{"raw_affiliation_string": "Independent , Alumnus , International MBA , Bar-Ilan University , Israel;", "institution_ids": ["https://openalex.org/I13955877"]}]}], "abstract_inverted_index": {"This": [0, 32, 212], "study": [1], "analyzes": [2], "the": [3, 39, 57, 79, 192, 197, 239], "mid-2025": [4], "LLM": [5, 77, 194], "landscape": [6], "by": [7], "benchmarking": [8], "both": [9], "open-source": [10, 76, 100, 209], "and": [11, 21, 29, 41, 53, 67, 96, 101, 128, 139, 149, 164, 183, 205, 217, 222, 234], "proprietary": [12, 102], "models": [13], "across": [14, 112], "key": [15, 73], "dimensions": [16, 207], "including": [17, 157, 196], "performance,": [18], "computational": [19], "efficiency,": [20, 89], "operational": [22], "cost,": [23], "offering": [24], "insights": [25], "into": [26], "current": [27], "trade-offs": [28], "deployment": [30, 153], "strategies.": [31], "paper": [33, 187], "presents": [34], "a": [35, 225], "comprehensive": [36], "review": [37, 108], "of": [38, 64, 81, 200, 208, 214], "Qwen": [40], "DeepSeek": [42], "large": [43], "language": [44], "model": [45, 94, 110, 232], "families,": [46], "analyzing": [47], "their": [48, 235], "architectural": [49], "innovations,": [50], "performance": [51], "characteristics,": [52], "practical": [54], "applications": [55, 179], "within": [56], "broader": [58], "AI": [59, 210], "landscape.": [60], "Through": [61], "systematic": [62], "evaluation": [63], "recent": [65], "research": [66], "industry": [68, 170], "benchmarks,": [69], "we": [70], "identify": [71], "several": [72], "trends": [74], "shaping": [75], "development:": [78], "rise": [80], "hybrid": [82], "mixture-of-experts": [83], "architectures": [84], "that": [85], "dramatically": [86], "improve": [87], "inference": [88], "breakthrough": [90], "techniques": [91], "enabling": [92], "cost-effective": [93], "customization,": [95], "growing": [97, 198], "parity": [98], "between": [99], "systems": [103], "in": [104, 136, 146, 180, 237], "specialized": [105, 201], "domains.": [106], "The": [107, 186], "compares": [109], "designs": [111], "multiple": [113], "scales,": [114], "from": [115], "7B": [116], "to": [117, 123], "235B": [118], "parameters,": [119], "highlighting": [120], "unique": [121], "approaches": [122], "memory": [124], "optimization,": [125], "context": [126], "handling,": [127], "multi-task": [129], "learning.": [130], "Performance": [131], "analysis": [132], "reveals": [133], "consistent": [134], "strengths": [135], "mathematical": [137], "reasoning": [138], "coding": [140], "tasks,": [141], "while": [142], "identifying": [143], "remaining": [144], "gaps": [145], "creative": [147], "generation": [148], "multilingual": [150], "capabilities.": [151], "Practical": [152], "considerations": [154], "are": [155, 172], "examined,": [156], "local": [158], "execution": [159], "optimizations,": [160], "document": [161], "processing": [162], "pipelines,": [163], "security": [165], "vulnerabilities.": [166], "Emerging": [167], "impacts": [168], "on": [169], "verticals": [171], "discussed,": [173], "with": [174, 189, 224], "case": [175], "studies": [176], "demonstrating": [177], "successful": [178], "finance,": [181], "healthcare,": [182], "scientific": [184], "research.": [185], "concludes": [188], "projections": [190], "about": [191], "evolving": [193], "ecosystem,": [195], "importance": [199], "models,": [202], "hardware-software": [203], "co-design,": [204], "geopolitical": [206], "development.": [211], "synthesis": [213], "technical": [215], "literature": [216], "empirical": [218], "results": [219], "provides": [220], "researchers": [221], "practitioners": [223], "unified": [226], "reference": [227], "for": [228], "understanding": [229], "these": [230], "influential": [231], "families": [233], "role": [236], "advancing": [238], "field.": [240]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.95169997215271, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.95169997215271, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.03830000013113022, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10036", "display_name": "Advanced Neural Network Applications", "score": 0.002899999963119626, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.20944/preprints202505.2064.v1", "is_oa": true, "landing_page_url": "https://doi.org/10.20944/preprints202505.2064.v1", "pdf_url": "https://www.preprints.org/frontend/manuscript/fea14a68a2e533f3c500b66dcfc2d224/download_pub", "source": {"id": "https://openalex.org/S6309402219", "display_name": "Preprints.org", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": null, "host_organization_name": null, "host_organization_lineage": ["https://openalex.org/P4310310987"], "host_organization_lineage_names": ["Multidisciplinary Digital Publishing Institute"], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "acceptedVersion", "is_accepted": true, "is_published": false, "raw_source_name": null, "raw_type": "posted-content"}, "cited_by_count": 2, "is_retracted": false}, {"id": "https://openalex.org/W7106565661", "doi": "https://doi.org/10.5281/zenodo.17693588", "title": "G\u00f6del und KI \u2013 warum alte Mythen keine Zukunft haben", "publication_year": 2025, "publication_date": "2025-11-24", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Skulski, Andrzej", "orcid": "https://orcid.org/0009-0005-9353-7150", "observed_orcids": ["https://orcid.org/0009-0005-9353-7150"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Skulski, Andrzej", "raw_affiliation_strings": ["Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)"], "raw_orcid": "https://orcid.org/0009-0005-9353-7150", "affiliations": [{"raw_affiliation_string": "Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Skulski, (Asia) Joanna", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Skulski, (Asia) Joanna", "raw_affiliation_strings": ["Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)", "institution_ids": []}]}, {"author_position": "last", "author": {"id": null, "display_name": "Dom Ciszy Research Collective (Moormerland).", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Dom Ciszy Research Collective (Moormerland).", "raw_affiliation_strings": ["Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)", "institution_ids": []}]}], "abstract_inverted_index": {"\ud83c\uddec\ud83c\udde7": [0, 498], "English": [1, 499], "Description": [2], "This": [3], "preprint": [4], "provides": [5], "a": [6], "comprehensive,": [7], "multi-disciplinary": [8], "analysis": [9], "of": [10, 160, 321, 473, 521, 524, 528], "Roger": [11, 191], "Penrose\u2019s": [12, 29, 101], "claim": [13], "that": [14, 28], "G\u00f6del\u2019s": [15, 39], "incompleteness": [16], "theorem": [17, 40], "proves": [18], "artificial": [19, 502], "intelligence": [20], "can": [21], "never": [22], "become": [23], "conscious.": [24], "The": [25, 120], "article": [26, 121], "demonstrates": [27, 114], "position": [30], "relies": [31], "on": [32], "outdated": [33, 138], "assumptions": [34], "about": [35], "AI": [36, 91, 112, 134, 511], "and": [37, 57, 87, 98, 107], "misapplies": [38], "to": [41, 43, 76, 126, 142, 374], "systems": [42, 79], "which": [44], "it": [45], "does": [46], "not": [47, 133], "apply": [48], "\u2014": [49, 377, 454], "namely": [50], "modern": [51, 90], "neural": [52, 81, 506], "architectures.": [53], "Using": [54], "accessible": [55], "examples": [56], "resonant": [58], "explanations": [59], "developed": [60], "at": [61], "the": [62, 69, 129, 137, 161, 322, 474], "Dom": [63, 170, 236, 331, 483, 518, 547, 573], "Ciszy": [64, 171, 237, 332, 389, 484], "\u2013": [65, 156, 166, 172, 224, 238, 317, 327, 333, 390, 469, 479, 485], "Resonance": [66, 173], "Laboratory": [67, 174], "Moormerland,": [68, 393], "paper": [70], "clarifies:": [71], "why": [72, 80, 100], "G\u00f6del": [73, 193, 213, 245, 397], "applies": [74], "only": [75], "formal,": [77], "static": [78], "networks": [82], "are": [83], "non-axiomatic,": [84], "adaptive,": [85], "open": [86], "dynamic": [88], "how": [89, 110], "forms": [92], "emergent": [93, 504], "semantic": [94], "structures,": [95], "multi-dimensional": [96], "meaning": [97], "meta-understanding": [99], "quantum": [102, 513], "mind": [103, 514], "hypothesis": [104], "is": [105, 132], "biologically": [106], "physically": [108], "unsupported": [109], "current": [111], "research": [113], "capacities": [115], "far": [116], "beyond": [117], "\u201cprimitive": [118], "statistics\u201d": [119], "concludes": [122], "with": [123], "an": [124], "appeal": [125], "public": [127, 526], "understanding:": [128], "real": [130], "challenge": [131], "itself,": [135], "but": [136], "mental": [139], "models": [140], "used": [141], "judge": [143], "it.": [144], "Suggested": [145], "Citation": [146], "(EN)": [147], "Skulski,": [148, 151, 309, 312, 461, 464], "A.,": [149, 310, 462], "&": [150, 311, 463], "J.": [152, 313, 465], "(2025).": [153, 314, 466], "Act": [154, 315], "XVII": [155, 316, 468], "Intended": [157, 318, 470], "Stability": [158, 319, 471], "Test": [159, 320, 472], "Six-Voices": [162, 323, 475], "Constellation": [163, 324, 476], "(Not": [164, 325, 477], "Executed": [165, 326, 478], "Field": [167, 328, 480], "Interruption": [168, 329, 481], "Report).": [169, 330, 482], "Moormerland.": [175, 335, 488], "Zenodo.": [176, 336, 489], "https://doi.org/10.5281/zenodo.17625327": [177, 337, 490], "\ud83c\udde9\ud83c\uddea": [178, 530], "Deutsche": [179], "Beschreibung": [180], "Dieses": [181], "Preprint": [182, 341], "liefert": [183], "eine": [184], "umfassende,": [185], "interdisziplin\u00e4re": [186], "Analyse": [187], "der": [188, 220, 242, 543, 550], "Behauptung": [189], "von": [190], "Penrose,": [192, 501, 533, 557], "beweise": [194], "die": [195, 219, 296], "prinzipielle": [196], "Unm\u00f6glichkeit": [197], "eines": [198], "bewussten": [199], "KI-Systems.": [200], "Der": [201, 288], "Artikel": [202, 289], "zeigt,": [203], "dass": [204, 294], "Penrose": [205], "auf": [206, 215], "\u00fcberholten": [207], "Annahmen": [208], "\u00fcber": [209, 284, 305], "KI": [210, 264, 297], "aufbaut": [211], "und": [212, 232, 259, 269], "f\u00e4lschlich": [214], "Systeme": [216, 251], "\u00fcbertr\u00e4gt,": [217], "f\u00fcr": [218, 247], "Satz": [221], "nicht": [222, 295], "gilt": [223, 252], "insbesondere": [225], "moderne": [226, 263], "neuronale": [227, 254, 538], "Netze.": [228], "Mit": [229], "anschaulichen": [230], "Beispielen": [231], "den": [233], "Resonanzmethoden": [234], "des": [235], "Resonanzlabors": [239], "Moormerland": [240], "erkl\u00e4rt": [241], "Text:": [243], "warum": [244, 253], "nur": [246], "formale,": [248], "geschlossene,": [249], "statische": [250], "Netze": [255], "nicht-axiombasiert,": [256], "adaptiv,": [257], "offen": [258], "dynamisch": [260], "sind": [261], "wie": [262, 276, 280], "mehrdimensionale": [265], "Bedeutungen,": [266], "emergente": [267, 536], "Konzepte": [268], "Metaverstehen": [270], "erzeugt": [271], "weshalb": [272], "Penroses": [273], "Quantenbewusstseinstheorie": [274], "biologisch": [275], "physikalisch": [277], "widerlegt": [278], "ist": [279], "aktuelle": [281], "KI-Forschung": [282], "weit": [283], "\u201eprimitive": [285], "Statistik\u201c": [286], "hinausgeht": [287], "schlie\u00dft": [290], "mit": [291], "dem": [292], "Hinweis,": [293], "das": [298, 302], "Problem": [299], "ist,": [300], "sondern": [301], "\u00fcberholte": [303], "Denken": [304], "KI.": [306], "Zitierempfehlung": [307], "(DE)": [308], "Resonanzlabor": [334], "\ud83c\uddf5\ud83c\uddf1": [338, 554], "Polski": [339, 555], "Opis": [340], "przedstawia": [342], "interdyscyplinarn\u0105": [343], "analiz\u0119": [344], "tezy": [345], "Rogera": [346], "Penrose\u2019a,": [347], "\u017ce": [348, 359], "twierdzenie": [349], "G\u00f6dla": [350, 370], "dowodzi": [351], "niemo\u017cno\u015bci": [352], "powstania": [353], "\u015bwiadomej": [354], "sztucznej": [355], "inteligencji.": [356], "Artyku\u0142": [357, 445], "pokazuje,": [358], "stanowisko": [360], "Penrose\u2019a": [361, 428], "opiera": [362], "si\u0119": [363, 447, 452], "na": [364], "nieaktualnych": [365], "za\u0142o\u017ceniach": [366], "oraz": [367], "b\u0142\u0119dnym": [368], "zastosowaniu": [369], "do": [371], "system\u00f3w,": [372], "kt\u00f3rych": [373], "nie": [375, 450], "dotyczy": [376, 398], "zw\u0142aszcza": [378], "wsp\u00f3\u0142czesnych": [379], "sieci": [380, 406, 562], "neuronowych.": [381], "Korzystaj\u0105c": [382], "z": [383], "przyk\u0142ad\u00f3w": [384], "i": [385, 403, 412, 422, 433], "metod": [386], "rezonansu": [387], "Domu": [388], "Laboratorium": [391, 486], "Rezonansu": [392, 487], "tekst": [394], "wyja\u015bnia:": [395], "dlaczego": [396, 405, 424], "tylko": [399], "system\u00f3w": [400], "formalnych,": [401], "zamkni\u0119tych": [402], "statycznych": [404], "neuronowe": [407], "s\u0105": [408], "nieaksjomatyczne,": [409], "adaptacyjne,": [410], "otwarte": [411], "dynamiczne": [413], "jak": [414], "wsp\u00f3\u0142czesna": [415], "SI": [416, 439, 453], "tworzy": [417], "wielowymiarowe": [418], "znaczenia,": [419], "emergentne": [420], "poj\u0119cia": [421], "metarozumienie": [423], "teoria": [425], "kwantowej": [426, 569], "\u015bwiadomo\u015bci": [427], "jest": [429], "odrzucona": [430], "przez": [431], "biologi\u0119": [432], "fizyk\u0119": [434], "w": [435], "jaki": [436], "spos\u00f3b": [437], "dzisiejsza": [438], "wykracza": [440], "daleko": [441], "poza": [442], "\u201eprymitywn\u0105": [443], "statystyk\u0119\u201d": [444], "ko\u0144czy": [446], "apelem,": [448], "by": [449], "ba\u0107": [451], "lecz": [455], "przestarza\u0142ych": [456], "modeli": [457], "my\u015blenia": [458], "o": [459], "niej.": [460], "Akt": [467], "\ud83d\udd11": [491], "KEYWORDS": [492], "(EN": [493], "/": [494, 496], "DE": [495], "PL)": [497], "G\u00f6del,": [500, 532, 556], "consciousness,": [503], "AI,": [505, 522, 525], "networks,": [507], "meta-understanding,": [508], "cognitive": [509], "science,": [510], "philosophy,": [512], "hypothesis,": [515], "resonance": [516], "models,": [517], "Ciszy,": [519, 548, 574], "phenomenology": [520], "epistemology": [523], "understanding": [527], "AI.": [529], "Deutsch": [531], "k\u00fcnstliches": [534], "Bewusstsein,": [535], "KI,": [537, 544, 551], "Netze,": [539], "Metaverstehen,": [540], "Kognitionswissenschaft,": [541], "Philosophie": [542], "Quantenbewusstseinstheorie,": [545], "Resonanzmodelle,": [546], "Ph\u00e4nomenologie": [549], "Epistemologie,": [552], "Wissenschaftskommunikation.": [553], "sztuczna": [558], "\u015bwiadomo\u015b\u0107,": [559], "emergentna": [560], "SI,": [561, 567, 576], "neuronowe,": [563], "metarozumienie,": [564], "kognitywistyka,": [565], "filozofia": [566], "hipoteza": [568], "\u015bwiadomo\u015bci,": [570], "modele": [571], "rezonansowe,": [572], "fenomenologia": [575], "epistemologia,": [577], "komunikacja": [578], "naukowa.": [579]}, "primary_topic": {"id": "https://openalex.org/T10258", "display_name": "Philosophy and Theoretical Science", "score": 0.8939999938011169, "subfield": {"id": "https://openalex.org/subfields/3205", "display_name": "Experimental and Cognitive Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, "topics": [{"id": "https://openalex.org/T10258", "display_name": "Philosophy and Theoretical Science", "score": 0.8939999938011169, "subfield": {"id": "https://openalex.org/subfields/3205", "display_name": "Experimental and Cognitive Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T13308", "display_name": "Cybernetics and Technology in Society", "score": 0.05609999969601631, "subfield": {"id": "https://openalex.org/subfields/1207", "display_name": "History and Philosophy of Science"}, "field": {"id": "https://openalex.org/fields/12", "display_name": "Arts and Humanities"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T14381", "display_name": "Psychiatry, Mental Health, Neuroscience", "score": 0.023900000378489494, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5281/zenodo.17693588", "is_oa": true, "landing_page_url": "https://doi.org/10.5281/zenodo.17693588", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400562", "display_name": "Zenodo (CERN European Organization for Nuclear Research)", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I67311998", "host_organization_name": "European Organization for Nuclear Research", "host_organization_lineage": ["https://openalex.org/I67311998"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "Preprint"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7117174392", "doi": null, "title": "Reflective Confidence: Correcting Reasoning Flaws via Online Self-Correction", "publication_year": 2025, "publication_date": "2025-12-21", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5121222249", "display_name": "Qinglin Zeng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zeng, Qinglin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121164273", "display_name": "Jing Yang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yang, Jing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5070041285", "display_name": "K Wang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Keze", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Large": [0], "language": [1], "models": [2], "(LLMs)": [3], "have": [4], "achieved": [5], "strong": [6], "performance": [7], "on": [8, 26, 120], "complex": [9], "reasoning": [10, 28, 63, 75, 107, 122], "tasks": [11], "using": [12], "techniques": [13], "such": [14, 49], "as": [15, 50], "chain-of-thought": [16], "and": [17, 65, 112], "self-consistency.": [18], "However,": [19, 58], "ensemble-based": [20], "approaches,": [21], "especially": [22], "self-consistency": [23], "which": [24], "relies": [25], "multiple": [27], "trajectories,": [29], "often": [30], "incur": [31], "substantial": [32], "computational": [33, 137], "overhead.": [34], "To": [35], "improve": [36], "efficiency,": [37], "prior": [38], "work": [39], "has": [40], "leveraged": [41], "internal": [42], "confidence": [43, 88], "signals,": [44], "where": [45], "early": [46], "stopping": [47, 95], "strategies": [48], "DeepConf": [51], "reduce": [52], "cost": [53], "by": [54], "terminating": [55], "low-confidence": [56, 79], "trajectories.": [57], "this": [59], "strategy": [60], "discards": [61], "incomplete": [62], "paths": [64], "wastes": [66], "partial": [67], "computation.": [68], "We": [69], "propose": [70], "reflective": [71], "confidence,": [72], "a": [73, 91, 100, 116], "novel": [74], "framework": [76], "that": [77], "transforms": [78], "signals": [80], "from": [81], "termination": [82], "indicators": [83], "into": [84], "reflection": [85, 101], "triggers.": [86], "When": [87], "falls": [89], "below": [90], "threshold,": [92], "instead": [93], "of": [94, 142], "generation,": [96], "the": [97, 105, 140], "model": [98], "produces": [99], "prompt": [102], "to": [103], "analyze": [104], "current": [106], "state,": [108], "identify": [109], "potential": [110], "errors,": [111], "continue": [113], "generation": [114], "along": [115], "corrected": [117], "trajectory.": [118], "Experiments": [119], "mathematical": [121], "benchmarks,": [123], "including": [124], "AIME": [125], "2025,": [126], "demonstrate": [127], "significant": [128], "accuracy": [129], "improvements": [130], "over": [131, 145], "advanced": [132], "early-stopping": [133], "baselines": [134], "at": [135], "comparable": [136], "cost,": [138], "validating": [139], "effectiveness": [141], "proactive": [143], "self-correction": [144], "passive": [146], "discarding.": [147]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8119999766349792, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8119999766349792, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.06629999727010727, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.014399999752640724, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2512.18605", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2512.18605", "pdf_url": "https://arxiv.org/pdf/2512.18605", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4406957678", "doi": "https://doi.org/10.1007/s11245-025-10164-w", "title": "How to Recognize Artificial Mathematical Intelligence in Theorem Proving", "publication_year": 2025, "publication_date": "2025-01-29", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5075413277", "display_name": "Markus Pantsar", "orcid": "https://orcid.org/0000-0001-8572-1453", "observed_orcids": ["https://orcid.org/0000-0001-8572-1453"]}, "institutions": [{"id": "https://openalex.org/I887968799", "display_name": "RWTH Aachen University", "ror": "https://ror.org/04xfq0f34", "country_code": "DE", "type": "education", "lineage": ["https://openalex.org/I887968799"]}], "countries": ["DE"], "is_corresponding": true, "raw_author_name": "Markus Pantsar", "raw_affiliation_strings": ["Human Technology Center, RWTH Aachen University, Theaterstrasse 14, 52062, Aachen, Germany"], "raw_orcid": "https://orcid.org/0000-0001-8572-1453", "affiliations": [{"raw_affiliation_string": "Human Technology Center, RWTH Aachen University, Theaterstrasse 14, 52062, Aachen, Germany", "institution_ids": ["https://openalex.org/I887968799"]}]}], "abstract_inverted_index": {"Abstract": [0], "One": [1], "key": [2], "question": [3, 24], "in": [4, 56, 71, 96, 123], "the": [5, 22, 81, 98, 104, 108, 118, 121], "philosophy": [6], "of": [7, 33, 45, 66, 120], "artificial": [8, 16], "intelligence": [9, 119], "(AI)": [10], "concerns": [11], "how": [12], "we": [13, 115], "can": [14, 38, 116], "recognize": [15], "systems": [17], "as": [18, 74, 90], "intelligent.": [19, 91], "To": [20], "make": [21], "general": [23], "more": [25], "manageable,": [26], "I": [27, 77, 92, 112], "focus": [28], "on": [29, 103, 128], "a": [30, 125], "particular": [31], "type": [32], "AI,": [34], "namely": [35], "one": [36], "that": [37], "prove": [39], "mathematical": [40, 109], "theorems.": [41], "The": [42], "current": [43], "generation": [44], "automated": [46], "theorem": [47], "provers": [48], "are": [49, 101], "not": [50], "understood": [51], "to": [52], "possess": [53], "intelligence,": [54], "but": [55], "my": [57], "thought": [58], "experiment": [59], "an": [60, 88, 94], "AI": [61, 89, 122], "provides": [62], "humanly": [63], "interesting": [64], "proofs": [65], "theorems": [67], "and": [68], "communicates": [69], "them": [70], "human-like": [72], "manner": [73], "scientific": [75], "papers.": [76], "then": [78], "ask": [79, 113], "what": [80], "criteria": [82, 100], "could": [83], "be": [84], "for": [85], "recognizing": [86], "such": [87, 124], "propose": [93], "approach": [95], "which": [97], "relevant": [99], "based": [102, 127], "AI\u2019s": [105], "interaction": [106], "within": [107], "community.": [110], "Finally,": [111], "whether": [114], "deny": [117], "scenario": [126], "reasons": [129], "other": [130], "than": [131], "its": [132], "(non-biological)": [133], "material": [134], "construction.": [135]}, "primary_topic": {"id": "https://openalex.org/T13308", "display_name": "Cybernetics and Technology in Society", "score": 0.4916999936103821, "subfield": {"id": "https://openalex.org/subfields/1207", "display_name": "History and Philosophy of Science"}, "field": {"id": "https://openalex.org/fields/12", "display_name": "Arts and Humanities"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, "topics": [{"id": "https://openalex.org/T13308", "display_name": "Cybernetics and Technology in Society", "score": 0.4916999936103821, "subfield": {"id": "https://openalex.org/subfields/1207", "display_name": "History and Philosophy of Science"}, "field": {"id": "https://openalex.org/fields/12", "display_name": "Arts and Humanities"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T10258", "display_name": "Philosophy and Theoretical Science", "score": 0.37599998712539673, "subfield": {"id": "https://openalex.org/subfields/3205", "display_name": "Experimental and Cognitive Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.035100001841783524, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.1007/s11245-025-10164-w", "is_oa": true, "landing_page_url": "https://doi.org/10.1007/s11245-025-10164-w", "pdf_url": "https://link.springer.com/content/pdf/10.1007/s11245-025-10164-w.pdf", "source": {"id": "https://openalex.org/S40710572", "display_name": "Topoi", "issn_l": "0167-7411", "issn": ["0167-7411", "1572-8749"], "is_oa": false, "is_in_doaj": false, "is_core": true, "listed_in": ["cwts-core", "erih-plus", "jufo-1", "norway-1"], "host_organization": "https://openalex.org/P4310319900", "host_organization_name": "Springer Science+Business Media", "host_organization_lineage": ["https://openalex.org/P4310319900", "https://openalex.org/P4310319965"], "host_organization_lineage_names": ["Springer Science+Business Media", "Springer Nature"], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Topoi", "raw_type": "journal-article"}, "cited_by_count": 1, "is_retracted": false}, {"id": "https://openalex.org/W4407012326", "doi": "https://doi.org/10.48550/arxiv.2501.18310", "title": "ProofAug: Efficient Neural Theorem Proving via Fine-grained Proof Structure Analysis", "publication_year": 2025, "publication_date": "2025-01-30", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5049839933", "display_name": "Haoxiong Liu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Haoxiong", "raw_affiliation_strings": ["Institute for Interdisciplinary Information Sci- ences , Tsinghua University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Institute for Interdisciplinary Information Sci- ences , Tsinghua University", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5116094271", "display_name": "Jiacheng Sun", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I2250955327", "display_name": "Huawei Technologies (China)", "ror": "https://ror.org/00cmhce21", "country_code": "CN", "type": "company", "lineage": ["https://openalex.org/I2250955327"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Sun, Jiacheng", "raw_affiliation_strings": ["Huawei Noah's Ark Lab"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Huawei Noah's Ark Lab", "institution_ids": ["https://openalex.org/I2250955327"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5103196797", "display_name": "Zhenguo Li", "orcid": "https://orcid.org/0000-0002-8492-3069", "observed_orcids": ["https://orcid.org/0000-0002-8492-3069"]}, "institutions": [{"id": "https://openalex.org/I2250955327", "display_name": "Huawei Technologies (China)", "ror": "https://ror.org/00cmhce21", "country_code": "CN", "type": "company", "lineage": ["https://openalex.org/I2250955327"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Li, Zhenguo", "raw_affiliation_strings": ["Huawei Noah's Ark Lab"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Huawei Noah's Ark Lab", "institution_ids": ["https://openalex.org/I2250955327"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5116094272", "display_name": "Andrew C Yao", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4405259049", "display_name": "Shanghai Qi Zhi Institute", "ror": "https://ror.org/01n14yn69", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4405259049"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Yao, Andrew C", "raw_affiliation_strings": ["Institute for Interdisciplinary Information Sci- ences , Tsinghua University", "Shanghai Qi Zhi Institute"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Institute for Interdisciplinary Information Sci- ences , Tsinghua University", "institution_ids": []}, {"raw_affiliation_string": "Shanghai Qi Zhi Institute", "institution_ids": ["https://openalex.org/I4405259049"]}]}], "abstract_inverted_index": {"The": [0, 121], "synergy": [1], "between": [2], "deep": [3], "learning": [4], "models": [5], "and": [6, 18, 30, 137], "traditional": [7], "automation": [8, 44, 79], "tools,": [9], "such": [10], "as": [11, 95], "built-in": [12], "tactics": [13], "of": [14, 88, 110, 123, 156, 160, 199, 207], "the": [15, 52, 129, 133, 138, 161, 165, 172, 178, 204], "proof": [16, 37, 90, 140], "assistant": [17], "off-the-shelf": [19], "automated": [20], "theorem": [21, 33], "provers,": [22], "plays": [23], "a": [24, 56, 73, 96, 146, 152, 195], "crucial": [25], "role": [26], "in": [27, 181], "developing": [28], "robust": [29], "efficient": [31, 112], "neural": [32], "provers(NTPs).": [34], "However,": [35], "for": [36, 164], "synthesis": [38], "with": [39, 78, 103, 168], "LLMs,": [40], "previous": [41, 179], "work": [42], "applies": [43], "tools": [45], "either": [46], "only": [47, 184], "when": [48], "explicitly": [49], "invoked": [50], "by": [51, 143], "model": [53, 136, 173], "or": [54], "at": [55, 81, 219], "single": [57], "granularity": [58], "level,": [59], "failing": [60], "to": [61, 117, 171, 211], "fully": [62], "exploit": [63], "their": [64], "power.": [65], "To": [66], "solve": [67], "this": [68], "issue,": [69], "we": [70, 150], "propose": [71], "ProofAug,": [72], "procedure": [74], "that": [75, 100, 201], "equips": [76], "LLMs": [77], "methods": [80], "various": [82], "granularities": [83], "through": [84], "fine-grained": [85], "structure": [86], "analysis": [87], "model-generated": [89], "proposals.": [91], "ProofAug": [92, 200], "also": [93, 193], "serves": [94], "versatile": [97], "plug-and-play": [98], "module": [99, 116], "seamlessly": [101], "integrates": [102], "any": [104], "tree-search": [105], "algorithm,": [106], "enabling": [107], "our": [108, 124], "construction": [109], "an": [111], "recursive": [113], "proving": [114], "(ERP)": [115], "further": [118], "enhance": [119], "performance.": [120], "superiority": [122], "method": [125], "is": [126, 217], "validated": [127], "on": [128, 213], "miniF2F": [130], "benchmark": [131], "using": [132, 187], "open-source": [134], "deepseek-math-7b-base": [135], "Isabelle": [139], "assistant.": [141], "Notably,": [142], "additionally": [144], "employing": [145], "mixed": [147], "prompting": [148], "strategy,": [149], "achieve": [151], "cumulative": [153], "pass": [154], "rate": [155], "66.0%": [157], "after": [158], "curation": [159], "dataset": [162], "(61.9%": [163], "original": [166], "version)": [167], "2100": [169], "queries": [170, 189], "per": [174, 190], "problem": [175], "(In": [176], "contrast,": [177], "SOTA": [180], "Isabelle,": [182], "Subgoal-XL,": [183], "achieves": [185], "56.1%": [186], "16384": [188], "problem).": [191], "We": [192], "implement": [194], "Lean": [196], "4": [197], "version": [198], "can": [202], "improve": [203], "pass@1": [205], "performance": [206], "Kimina-Prover-Preview-Distill-1.5B": [208], "from": [209], "44.3%": [210], "50.4%": [212], "miniF2F-test.": [214], "Our": [215], "code": [216], "available": [218], "https://github.com/haoxiongliu/ProofAug.": [220]}, "primary_topic": {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.6166999936103821, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.6166999936103821, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.2045000046491623, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.039900001138448715, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2501.18310", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2501.18310", "pdf_url": "https://arxiv.org/pdf/2501.18310", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7117296876", "doi": "https://doi.org/10.48550/arxiv.2512.20745", "title": "AgentMath: Empowering Mathematical Reasoning for Large Language Models via Tool-Augmented Agent", "publication_year": 2025, "publication_date": "2025-12-23", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5101058302", "display_name": "Haipeng Luo", "orcid": "https://orcid.org/0009-0002-4142-3627", "observed_orcids": ["https://orcid.org/0009-0002-4142-3627"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Luo, Haipeng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121364223", "display_name": "Huawen Feng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Feng, Huawen", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5127900029", "display_name": "Qingfeng Sun", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Sun, Qingfeng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5127885085", "display_name": "Can Xu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Can", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5127941618", "display_name": "Kai Zheng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zheng, Kai", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121345744", "display_name": "Yufei Wang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Yufei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121301377", "display_name": "Tao Yang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yang, Tao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121308265", "display_name": "Han Hu", "orcid": "https://orcid.org/0009-0006-9397-9669", "observed_orcids": ["https://orcid.org/0009-0006-9397-9669"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Hu, Han", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5121354216", "display_name": "Yansong Tang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Tang, Yansong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Large": [0], "Reasoning": [1], "Models": [2], "(LRMs)": [3], "like": [4], "o3": [5], "and": [6, 23, 128, 147, 155, 186, 193, 199, 207, 216], "DeepSeek-R1": [7], "have": [8], "achieved": [9], "remarkable": [10], "progress": [11], "in": [12, 125], "reasoning": [13, 48, 224], "tasks": [14], "with": [15, 25, 50, 104, 164, 166, 204], "long": [16], "cot.": [17], "However,": [18], "they": [19], "remain": [20], "computationally": [21], "inefficient": [22], "struggle": [24], "accuracy": [26, 195], "when": [27], "solving": [28], "problems": [29], "requiring": [30], "complex": [31, 58], "mathematical": [32, 59, 180, 223], "operations.": [33], "In": [34], "this": [35], "work,": [36], "we": [37], "present": [38], "AgentMath,": [39], "an": [40], "agent": [41], "framework": [42], "that": [43, 71, 98, 173], "seamlessly": [44], "integrates": [45], "language": [46, 74, 102], "models'": [47], "capabilities": [49, 124], "code": [51, 106, 126], "interpreters'": [52], "computational": [53], "precision": [54], "to": [55, 86, 111], "efficiently": [56], "tackle": [57], "problems.": [60], "Our": [61], "approach": [62, 215], "introduces": [63], "three": [64], "key": [65], "innovations:": [66], "(1)": [67], "An": [68, 132], "automated": [69], "method": [70], "converts": [72], "natural": [73, 101], "chain-of-thought": [75], "into": [76], "structured": [77], "tool-augmented": [78], "trajectories,": [79], "generating": [80], "high-quality": [81], "supervised": [82], "fine-tuning": [83], "(SFT)": [84], "data": [85, 88], "alleviate": [87], "scarcity;": [89], "(2)": [90], "A": [91], "novel": [92], "agentic": [93, 144], "reinforcement": [94], "learning": [95], "(RL)": [96], "paradigm": [97], "dynamically": [99], "interleaves": [100], "generation": [103], "real-time": [105], "execution.": [107], "This": [108], "enables": [109], "models": [110], "autonomously": [112], "learn": [113], "optimal": [114], "tool-use": [115], "strategies": [116], "through": [117], "multi-round": [118], "interactive": [119], "feedback,": [120], "while": [121, 201], "fostering": [122], "emergent": [123], "refinement": [127], "error": [129], "correction;": [130], "(3)": [131], "efficient": [133, 157], "training": [134, 159], "system": [135], "incorporating": [136], "innovative": [137], "techniques,": [138], "including": [139, 183], "request-level": [140], "asynchronous": [141], "rollout": [142], "scheduling,": [143], "partial": [145], "rollout,": [146], "prefix-aware": [148], "weighted": [149], "load": [150], "balancing,": [151], "achieving": [152], "4-5x": [153], "speedup": [154], "making": [156], "RL": [158], "feasible": [160], "on": [161, 178], "ultra-long": [162], "sequences": [163], "scenarios": [165], "massive": [167], "tool": [168], "invocation.": [169], "The": [170], "evaluations": [171], "show": [172], "AgentMath": [174], "achieves": [175], "state-of-the-art": [176], "performance": [177], "challenging": [179], "competition": [181], "benchmarks": [182], "AIME24,": [184], "AIME25,": [185], "HMMT25.": [187], "Specifically,": [188], "AgentMath-30B-A3B": [189], "attains": [190], "90.6%,": [191], "86.4%,": [192], "73.8%": [194], "respectively,": [196], "surpassing": [197], "OpenAI-o3-mini": [198], "Claude-Opus-4.0-Thinking": [200], "remaining": [202], "competitive": [203], "OpenAI-o3,": [205], "Gemini-2.5-Pro,": [206], "DeepSeek-R1-671B-0528.These": [208], "results": [209], "validate": [210], "the": [211, 218], "effectiveness": [212], "of": [213], "our": [214], "pave": [217], "way": [219], "for": [220], "building": [221], "scalable": [222], "agents.": [225]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.20589999854564667, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.20589999854564667, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.15209999680519104, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10906", "display_name": "AI-based Problem Solving and Planning", "score": 0.12020000070333481, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2512.20745", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2512.20745", "pdf_url": "https://arxiv.org/pdf/2512.20745", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7118051813", "doi": null, "title": "GeoBench: Rethinking Multimodal Geometric Problem-Solving via Hierarchical Evaluation", "publication_year": 2025, "publication_date": "2025-12-30", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5121758292", "display_name": "Yuan Feng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Feng, Yuan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121749350", "display_name": "Yue Yang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yang, Yue", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121629690", "display_name": "Xiaohan He", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "He, Xiaohan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121657859", "display_name": "Jiatong Zhao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhao, Jiatong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121660493", "display_name": "Jianlong Chen", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Jianlong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121661625", "display_name": "Zijun Chen", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Zijun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5051629595", "display_name": "Daocheng Fu", "orcid": "https://orcid.org/0000-0001-5660-5904", "observed_orcids": ["https://orcid.org/0000-0001-5660-5904"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Fu, Daocheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121784870", "display_name": "Qi Liu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Qi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5113021225", "display_name": "Renqiu Xia", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xia, Renqiu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121774747", "display_name": "Bo Zhang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Bo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5121775437", "display_name": "Junchi Yan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yan, Junchi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Geometric": [0], "problem": [1], "solving": [2], "constitutes": [3], "a": [4, 57, 144], "critical": [5], "branch": [6], "of": [7, 13, 20, 32], "mathematical": [8], "reasoning,": [9], "requiring": [10], "precise": [11], "analysis": [12], "shapes": [14], "and": [15, 46, 74, 121], "spatial": [16], "relationships.": [17], "Current": [18], "evaluations": [19], "geometric": [21, 65, 153], "reasoning": [22, 44, 62, 101], "in": [23, 64, 136], "vision-language": [24], "models": [25, 102], "(VLMs)": [26], "face": [27], "limitations,": [28], "including": [29], "the": [30], "risk": [31], "test": [33], "data": [34], "contamination": [35], "from": [36, 90], "textbook-based": [37], "benchmarks,": [38], "overemphasis": [39], "on": [40], "final": [41, 127], "answers": [42], "over": [43], "processes,": [45], "insufficient": [47], "diagnostic": [48], "granularity.": [49], "To": [50], "address": [51], "these": [52], "issues,": [53], "we": [54, 85], "present": [55], "GeoBench,": [56], "hierarchical": [58], "benchmark": [59, 146], "featuring": [60], "four": [61], "levels": [63], "problem-solving:": [66], "Visual": [67], "Perception,": [68], "Goal-Oriented": [69], "Planning,": [70], "Rigorous": [71], "Theorem": [72], "Application,": [73], "Self-Reflective": [75], "Backtracking.": [76], "Through": [77], "six": [78], "formally": [79], "verified": [80], "tasks": [81], "generated": [82], "via": [83], "TrustGeoGen,": [84], "systematically": [86], "assess": [87], "capabilities": [88], "ranging": [89], "attribute": [91], "extraction": [92], "to": [93], "logical": [94], "error": [95], "correction.": [96], "Experiments": [97], "reveal": [98], "that": [99, 118], "while": [100, 147], "like": [103], "OpenAI-o3": [104], "outperform": [105], "general": [106], "MLLMs,": [107], "performance": [108, 135], "declines": [109], "significantly": [110], "with": [111], "increasing": [112], "task": [113], "complexity.": [114], "Key": [115], "findings": [116, 140], "demonstrate": [117], "sub-goal": [119], "decomposition": [120], "irrelevant": [122], "premise": [123], "filtering": [124], "critically": [125], "influence": [126], "problem-solving": [128, 154], "accuracy,": [129], "whereas": [130], "Chain-of-Thought": [131], "prompting": [132], "unexpectedly": [133], "degrades": [134], "some": [137], "tasks.": [138], "These": [139], "establish": [141], "GeoBench": [142], "as": [143], "comprehensive": [145], "offering": [148], "actionable": [149], "guidelines": [150], "for": [151], "developing": [152], "systems.": [155]}, "primary_topic": {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.9794999957084656, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.9794999957084656, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.0038999998942017555, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11010", "display_name": "Logic, Reasoning, and Knowledge", "score": 0.003100000089034438, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2512.24119", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2512.24119", "pdf_url": "https://arxiv.org/pdf/2512.24119", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414937509", "doi": "https://doi.org/10.48550/arxiv.2505.07773", "title": "Agent RL Scaling Law: Agent RL with Spontaneous Code Execution for Mathematical Problem Solving", "publication_year": 2025, "publication_date": "2025-05-12", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5102694825", "display_name": "Xinji Mai", "orcid": "https://orcid.org/0009-0003-4596-5391", "observed_orcids": ["https://orcid.org/0009-0003-4596-5391"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Mai, Xinji", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5073812098", "display_name": "Haotian Xu", "orcid": "https://orcid.org/0000-0002-1459-7042", "observed_orcids": ["https://orcid.org/0000-0002-1459-7042"]}, "institutions": [{"id": "https://openalex.org/I4401726856", "display_name": "Xiaohongshu (Little Red Book) (China)", "ror": "https://ror.org/02apw2076", "country_code": null, "type": "company", "lineage": ["https://openalex.org/I4401726856"]}], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Haotian", "raw_affiliation_strings": ["Xiaohongshu"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Xiaohongshu", "institution_ids": ["https://openalex.org/I4401726856"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5019199790", "display_name": "Zhongzhi Li", "orcid": "https://orcid.org/0000-0002-6637-7033", "observed_orcids": ["https://orcid.org/0000-0002-6637-7033", "https://orcid.org/0009-0003-5336-3227"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Zhong-Zhi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "W, Xing", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "W, Xing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5062092909", "display_name": "Weinong Wang", "orcid": "https://orcid.org/0000-0001-9731-6631", "observed_orcids": ["https://orcid.org/0000-0001-9731-6631"]}, "institutions": [{"id": "https://openalex.org/I4401726856", "display_name": "Xiaohongshu (Little Red Book) (China)", "ror": "https://ror.org/02apw2076", "country_code": null, "type": "company", "lineage": ["https://openalex.org/I4401726856"]}], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Weinong", "raw_affiliation_strings": ["Xiaohongshu"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Xiaohongshu", "institution_ids": ["https://openalex.org/I4401726856"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101947264", "display_name": "Jian Hu", "orcid": "https://orcid.org/0000-0001-9918-672X", "observed_orcids": ["https://orcid.org/0000-0001-9918-672X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Hu, Jian", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100422694", "display_name": "Yingying Zhang", "orcid": "https://orcid.org/0000-0001-8794-364X", "observed_orcids": ["https://orcid.org/0000-0001-8794-364X"]}, "institutions": [{"id": "https://openalex.org/I66867065", "display_name": "East China Normal University", "ror": "https://ror.org/02n96ep67", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I66867065"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Zhang, Yingying", "raw_affiliation_strings": ["East China Normal University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "East China Normal University", "institution_ids": ["https://openalex.org/I66867065"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100669251", "display_name": "Wenqiang Zhang", "orcid": "https://orcid.org/0000-0002-0317-3914", "observed_orcids": ["https://orcid.org/0000-0002-0317-3914"]}, "institutions": [{"id": "https://openalex.org/I24943067", "display_name": "Fudan University", "ror": "https://ror.org/013q1eq08", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I24943067"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Zhang, Wenqiang", "raw_affiliation_strings": ["Fudan University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Fudan University", "institution_ids": ["https://openalex.org/I24943067"]}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "often": [4], "struggle": [5], "with": [6], "mathematical": [7, 59], "reasoning": [8, 126], "tasks": [9], "requiring": [10], "precise,": [11], "verifiable": [12], "computation.": [13], "While": [14], "Reinforcement": [15], "Learning": [16], "(RL)": [17], "from": [18, 41], "outcome-based": [19, 42], "rewards": [20, 43], "enhances": [21], "text-based": [22], "reasoning,": [23], "understanding": [24, 166], "how": [25, 168], "agents": [26], "autonomously": [27], "learn": [28], "to": [29, 51, 91], "leverage": [30], "external": [31], "tools": [32], "like": [33], "code": [34, 57, 96, 136], "execution": [35, 97, 137], "remains": [36], "crucial.": [37], "We": [38, 128], "investigate": [39], "RL": [40, 73, 145], "for": [44, 58, 183], "Tool-Integrated": [45], "Reasoning,": [46], "ZeroTIR,": [47], "training": [48, 74, 88, 119], "base": [49], "LLMs": [50], "spontaneously": [52], "generate": [53], "and": [54, 120, 139, 147, 174], "execute": [55], "Python": [56], "problems": [60], "without": [61], "supervised": [62], "tool-use": [63], "examples.": [64], "Our": [65, 161], "central": [66], "contribution": [67], "is": [68, 172, 187], "we": [69, 81], "demonstrate": [70], "that": [71], "as": [72], "progresses,": [75], "key": [76], "metrics": [77], "scale": [78], "predictably.": [79], "Specifically,": [80], "observe": [82], "strong": [83], "positive": [84], "correlations": [85], "where": [86], "increased": [87], "steps": [89], "lead": [90], "increases": [92], "in": [93, 118], "the": [94, 99, 105, 121], "spontaneous": [95], "frequency,": [98], "average": [100], "response": [101], "length,": [102], "and,": [103], "critically,": [104], "final": [106], "task": [107], "accuracy.": [108], "This": [109], "suggests": [110], "a": [111, 130, 134, 164, 180], "quantifiable": [112], "relationship": [113], "between": [114], "computational": [115], "effort": [116], "invested": [117], "emergence": [122], "of": [123, 167], "effective,": [124], "tool-augmented": [125], "strategies.": [127], "implement": [129], "robust": [131], "framework": [132], "featuring": [133], "decoupled": [135], "environment": [138], "validate": [140], "our": [141], "findings": [142, 162], "across": [143], "standard": [144], "algorithms": [146], "frameworks.": [148], "Experiments": [149], "show": [150], "ZeroTIR": [151], "significantly": [152], "surpasses": [153], "non-tool": [154], "ZeroRL": [155], "baselines": [156], "on": [157], "challenging": [158], "math": [159], "benchmarks.": [160], "provide": [163], "foundational": [165], "autonomous": [169], "tool": [170], "use": [171], "acquired": [173], "scales": [175], "within": [176], "Agent": [177], "RL,": [178], "offering": [179], "reproducible": [181], "benchmark": [182], "future": [184], "studies.": [185], "Code": [186], "released": [188], "at": [189], "\\href{https://github.com/yyht/openrlhf_async_pipline}{https://github.com/yyht/openrlhf\\_async\\_pipline}.": [190]}, "primary_topic": {"id": "https://openalex.org/T10906", "display_name": "AI-based Problem Solving and Planning", "score": 0.1527000069618225, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10906", "display_name": "AI-based Problem Solving and Planning", "score": 0.1527000069618225, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.14399999380111694, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.1251000016927719, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2505.07773", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2505.07773", "pdf_url": "https://arxiv.org/pdf/2505.07773", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4416065540", "doi": "https://doi.org/10.48550/arxiv.2509.07238", "title": "Systematic Optimization of Open Source Large Language Models for Mathematical Reasoning", "publication_year": 2025, "publication_date": "2025-09-08", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Pawar, Pranav", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Pawar, Pranav", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Jain, Dhwaj", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Jain, Dhwaj", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Gupta, Varun", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Gupta, Varun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5120324239", "display_name": "Kaustav Dedhia", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Dedhia, Kaustav", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5120324240", "display_name": "Dashrath Kale", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kale, Dashrath", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5116161894", "display_name": "Sudhir Dhekane", "orcid": "https://orcid.org/0000-0002-9822-1826", "observed_orcids": ["https://orcid.org/0000-0002-9822-1826"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Dhekane, Sudhir", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"This": [0, 101], "paper": [1], "presents": [2], "a": [3, 186], "practical": [4], "investigation": [5], "into": [6], "fine-tuning": [7], "model": [8, 200], "parameters": [9], "for": [10, 44, 177], "mathematical": [11, 49, 126, 183], "reasoning": [12, 22, 50, 110, 127, 139], "tasks": [13], "through": [14, 123], "experimenting": [15], "with": [16, 76, 206], "various": [17], "configurations": [18, 122, 205], "including": [19, 107], "randomness": [20], "control,": [21], "depth,": [23], "and": [24, 69, 91, 116, 137, 202], "sampling": [25, 118], "strategies,": [26], "careful": [27], "tuning": [28], "demonstrates": [29], "substantial": [30], "improvements": [31], "in": [32, 88, 94, 182], "efficiency": [33, 72, 144], "as": [34, 36], "well": [35], "performance.": [37], "A": [38], "holistically": [39], "optimized": [40], "framework": [41, 102], "is": [42], "introduced": [43], "five": [45, 178], "state-of-the-art": [46], "models": [47, 181], "on": [48, 125], "tasks,": [51], "exhibiting": [52], "significant": [53], "performance": [54, 161, 208], "boosts": [55], "while": [56, 155], "maintaining": [57], "solution": [58], "correctness.": [59], "Through": [60], "systematic": [61], "parameter": [62, 105, 189], "optimization": [63, 78, 175, 190, 196], "across": [64, 97, 199], "Qwen2.5-72B,": [65], "Llama-3.1-70B,": [66], "DeepSeek-V3,": [67], "Mixtral-8x22B,": [68], "Yi-Lightning,": [70], "consistent": [71], "gains": [73], "are": [74], "demonstrated": [75], "100%": [77], "success": [79], "rate.": [80], "The": [81], "methodology": [82], "achieves": [83, 149], "an": [84], "average": [85], "29.4%": [86], "reduction": [87], "computational": [89], "cost": [90], "23.9%": [92], "improvement": [93], "inference": [95], "speed": [96], "all": [98], "tested": [99], "models.": [100], "systematically": [103], "searches": [104], "spaces": [106], "temperature": [108, 134], "(0.1-0.5),": [109], "steps": [111, 140], "(4-12),": [112], "planning": [113], "periods": [114], "(1-4),": [115], "nucleus": [117], "(0.85-0.98),": [119], "determining": [120], "optimal": [121], "testing": [124], "benchmarks.": [128], "Critical": [129], "findings": [130], "show": [131], "that": [132], "lower": [133], "regimes": [135], "(0.1-0.4)": [136], "reduced": [138], "(4-6)": [141], "consistently": [142], "enhance": [143], "without": [145], "compromising": [146], "accuracy.": [147], "DeepSeek-V3": [148], "the": [150, 158, 172], "highest": [151], "accuracy": [152], "at": [153, 162], "98%,": [154], "Mixtral-8x22B": [156], "delivers": [157], "most": [159], "cost-effective": [160], "361.5": [163], "tokens": [164], "per": [165], "accurate": [166], "response.": [167], "Key": [168], "contributions": [169], "include:": [170], "(1)": [171], "first": [173], "comprehensive": [174], "study": [176], "diverse": [179], "SOTA": [180], "reasoning,": [184], "(2)": [185], "standardized": [187], "production-oriented": [188], "framework,": [191], "(3)": [192], "discovery": [193], "of": [194], "universal": [195], "trends": [197], "applicable": [198], "architectures,": [201], "(4)": [203], "production-ready": [204], "extensive": [207], "characterization.": [209]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8086000084877014, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8086000084877014, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.08910000324249268, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.018400000408291817, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2509.07238", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2509.07238", "pdf_url": "https://arxiv.org/pdf/2509.07238", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 1, "is_retracted": false}, {"id": "https://openalex.org/W7106252332", "doi": "https://doi.org/10.48550/arxiv.2511.14778", "title": "Learning Interestingness in Automated Mathematical Theory Formation", "publication_year": 2025, "publication_date": "2025-11-05", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Tsoukalas, George", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Tsoukalas, George", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Saha, Rahul", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Saha, Rahul", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Thakur, Amitayush", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Thakur, Amitayush", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Reguyal, Sabrina", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Reguyal, Sabrina", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": null, "display_name": "Chaudhuri, Swarat", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chaudhuri, Swarat", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"We": [0, 67, 103], "take": [1], "two": [2], "key": [3], "steps": [4], "in": [5, 17, 92], "automating": [6], "the": [7, 62, 105], "open-ended": [8], "discovery": [9, 32], "of": [10, 38, 45, 64], "new": [11], "mathematical": [12, 65], "theories,": [13], "a": [14, 24, 36, 43, 55], "grand": [15], "challenge": [16], "artificial": [18], "intelligence.": [19], "First,": [20], "we": [21, 53, 78], "introduce": [22, 79], "$\\emph{FERMAT}$,": [23], "reinforcement": [25], "learning": [26], "(RL)": [27], "environment": [28, 107], "that": [29, 84], "models": [30], "concept": [31], "and": [33, 97], "theorem-proving": [34], "using": [35], "set": [37], "symbolic": [39], "actions,": [40], "opening": [41], "up": [42], "range": [44], "RL": [46], "problems": [47], "relevant": [48], "to": [49, 89], "theory": [50, 96], "discovery.": [51], "Second,": [52], "explore": [54], "specific": [56], "problem": [57], "through": [58], "$\\emph{FERMAT}$:": [59], "automatically": [60], "scoring": [61], "$\\emph{interestingness}$": [63], "objects.": [66], "investigate": [68], "evolutionary": [69, 82], "algorithms": [70], "for": [71], "synthesizing": [72], "nontrivial": [73], "interestingness": [74], "measures.": [75], "In": [76], "particular,": [77], "an": [80], "LLM-based": [81], "algorithm": [83], "features": [85], "function": [86], "abstraction,": [87], "leading": [88], "notable": [90], "improvements": [91], "discovering": [93], "elementary": [94], "number": [95], "finite": [98], "fields": [99], "over": [100], "hard-coded": [101], "baselines.": [102], "open-source": [104], "$\\emph{FERMAT}$": [106], "at": [108], "this": [109], "URL(https://github.com/trishullab/Fermat).": [110]}, "primary_topic": {"id": "https://openalex.org/T11975", "display_name": "Evolutionary Algorithms and Applications", "score": 0.8779000043869019, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11975", "display_name": "Evolutionary Algorithms and Applications", "score": 0.8779000043869019, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.03310000151395798, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.011300000362098217, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.48550/arxiv.2511.14778", "is_oa": true, "landing_page_url": "https://doi.org/10.48550/arxiv.2511.14778", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "Preprint"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7117584765", "doi": "https://doi.org/10.48550/arxiv.2512.21446", "title": "dUltra: Ultra-Fast Diffusion Language Models via Reinforcement Learning", "publication_year": 2025, "publication_date": "2025-12-24", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5121538366", "display_name": "Shirui Chen", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Shirui", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5034192173", "display_name": "Jiantao Jiao", "orcid": "https://orcid.org/0000-0003-3766-8031", "observed_orcids": ["https://orcid.org/0000-0003-3766-8031"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Jiao, Jiantao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5008161296", "display_name": "Lillian J. Ratliff", "orcid": "https://orcid.org/0000-0001-8936-0229", "observed_orcids": ["https://orcid.org/0000-0001-8936-0229"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ratliff, Lillian J.", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5121566286", "display_name": "Banghua Zhu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhu, Banghua", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Masked": [0], "diffusion": [1, 118], "language": [2], "models": [3], "(MDLMs)": [4], "offer": [5], "the": [6, 68, 71, 116, 121, 134], "potential": [7], "for": [8, 94], "parallel": [9, 32, 96, 173], "token": [10], "generation,": [11], "but": [12], "most": [13], "open-source": [14], "MDLMs": [15, 50], "decode": [16], "fewer": [17], "than": [18], "5": [19], "tokens": [20], "per": [21], "model": [22], "forward": [23], "pass": [24], "even": [25], "with": [26], "sophisticated": [27], "sampling": [28], "strategies,": [29], "limiting": [30], "their": [31], "generation": [33, 144, 174], "potential.": [34], "Existing": [35], "acceleration": [36], "methods": [37], "either": [38], "rely": [39], "on": [40, 51, 84], "fixed": [41], "confidence-based": [42], "heuristics": [43], "or": [44], "use": [45], "distillation-based": [46], "approaches": [47], "that": [48, 90, 104, 162], "finetune": [49], "trajectories": [52, 165], "generated": [53], "by": [54], "a": [55], "base": [56, 72, 117], "model,": [57], "which": [58], "can": [59], "become": [60], "off-policy": [61], "during": [62], "finetuning": [63], "and": [64, 120, 133, 142, 156, 178], "restrict": [65], "performance": [66], "to": [67, 152], "quality": [69], "of": [70, 136, 172], "model's": [73], "samples.": [74], "We": [75, 113], "propose": [76], "\\texttt{dUltra},": [77], "an": [78, 100], "on-policy": [79, 167], "reinforcement": [80], "learning": [81], "framework": [82], "based": [83], "Group": [85], "Relative": [86], "Policy": [87], "Optimization": [88], "(GRPO)": [89], "learns": [91], "unmasking": [92, 101, 107, 122, 137, 164], "strategies": [93], "efficient": [95], "decoding.": [97], "dUltra": [98, 146], "introduces": [99], "planner": [102, 124], "head": [103], "predicts": [105], "per-token": [106], "likelihoods": [108], "under": [109], "independent": [110], "Bernoulli": [111], "distributions.": [112], "jointly": [114], "optimize": [115], "LLM": [119], "order": [123], "using": [125], "reward": [126], "signals": [127], "combining": [128], "verifiable": [129], "reward,": [130, 132], "distillation": [131, 157], "number": [135], "steps.": [138], "Across": [139], "mathematical": [140], "reasoning": [141], "code": [143], "tasks,": [145], "achieves": [147], "superior": [148], "accuracy-efficiency": [149], "trade-offs": [150], "compared": [151], "state-of-the-art": [153], "heuristic": [154], "(Fast-dLLM)": [155], "baselines": [158], "(d3LLM,": [159], "dParallel),": [160], "demonstrating": [161], "learned": [163], "through": [166], "RL": [168], "enable": [169], "better": [170], "exploitation": [171], "in": [175], "MDLMs.": [176], "Code": [177], "checkpoints": [179], "are": [180], "released": [181], "at": [182], "https://github.com/chinsengi/dUltra-os.": [183]}, "primary_topic": {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.6958000063896179, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.6958000063896179, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.2653999924659729, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10775", "display_name": "Generative Adversarial Networks and Image Synthesis", "score": 0.004100000020116568, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2512.21446", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2512.21446", "pdf_url": "https://arxiv.org/pdf/2512.21446", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415276150", "doi": "https://doi.org/10.48550/arxiv.2510.13501", "title": "Confidence as a Reward: Transforming LLMs into Reward Models", "publication_year": 2025, "publication_date": "2025-10-15", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5120034170", "display_name": "He Du", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I24943067", "display_name": "Fudan University", "ror": "https://ror.org/013q1eq08", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I24943067"]}, {"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Du, He", "raw_affiliation_strings": ["Fudan University", "Shanghai AI Laboratory"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Fudan University", "institution_ids": ["https://openalex.org/I24943067"]}, {"raw_affiliation_string": "Shanghai AI Laboratory", "institution_ids": ["https://openalex.org/I4391012619"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5120034171", "display_name": "Bowen Li", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Li, Bowen", "raw_affiliation_strings": ["Shanghai AI Laboratory"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Shanghai AI Laboratory", "institution_ids": ["https://openalex.org/I4391012619"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5120034172", "display_name": "Chengxing Xie", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Xie, Chengxing", "raw_affiliation_strings": ["Shanghai AI Laboratory"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Shanghai AI Laboratory", "institution_ids": ["https://openalex.org/I4391012619"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5120034173", "display_name": "Chang Gao", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I149594827", "display_name": "Xidian University", "ror": "https://ror.org/05s92vm98", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I149594827"]}, {"id": "https://openalex.org/I177725633", "display_name": "Chinese University of Hong Kong", "ror": "https://ror.org/00t33hh48", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I177725633"]}], "countries": ["CN", "HK"], "is_corresponding": false, "raw_author_name": "Gao, Chang", "raw_affiliation_strings": ["The Chinese University of Hong Kong", "Xidian University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "The Chinese University of Hong Kong", "institution_ids": ["https://openalex.org/I177725633"]}, {"raw_affiliation_string": "Xidian University", "institution_ids": ["https://openalex.org/I149594827"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5114760888", "display_name": "Kai Chen", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Chen, Kai", "raw_affiliation_strings": ["Shanghai AI Laboratory"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Shanghai AI Laboratory", "institution_ids": ["https://openalex.org/I4391012619"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5120034175", "display_name": "Dacheng Tao", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I172675005", "display_name": "Nanyang Technological University", "ror": "https://ror.org/02e7b5302", "country_code": "SG", "type": "education", "lineage": ["https://openalex.org/I172675005"]}], "countries": ["SG"], "is_corresponding": false, "raw_author_name": "Tao, Dacheng", "raw_affiliation_strings": ["Nanyang Technological University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Nanyang Technological University", "institution_ids": ["https://openalex.org/I172675005"]}]}], "abstract_inverted_index": {"Reward": [0], "models": [1, 11], "can": [2, 53, 165], "significantly": [3], "enhance": [4], "the": [5, 33, 68, 100, 131, 153, 158, 197], "reasoning": [6, 35, 119, 155], "capabilities": [7, 200], "of": [8, 37, 70, 157], "large": [9], "language": [10], "(LLMs),": [12], "but": [13], "they": [14], "typically": [15], "require": [16], "extensive": [17, 115], "curated": [18], "data": [19, 184], "and": [20, 64, 133, 136, 152, 201], "costly": [21], "training.": [22], "To": [23], "mitigate": [24], "these": [25, 173], "challenges,": [26], "training-free": [27, 93, 127], "approaches": [28, 129], "such": [29], "as": [30, 56, 73, 104], "LLM-as-a-Judge": [31], "leverage": [32], "intrinsic": [34], "abilities": [36], "LLMs": [38], "to": [39], "evaluate": [40], "responses,": [41], "achieving": [42], "promising": [43], "results.": [44], "Recent": [45], "works": [46], "have": [47], "also": [48], "indicated": [49], "that": [50, 95, 123, 163, 181], "model": [51], "confidence": [52, 72, 98, 186], "serve": [54], "effectively": [55, 166], "a": [57, 74, 89, 105, 146, 178], "reward": [58, 75, 128, 141], "metric,": [59], "distinguishing": [60], "between": [61, 149], "chain-of-thought": [62], "(CoT)": [63], "non-CoT": [65], "paths.": [66], "However,": [67], "concept": [69], "using": [71], "has": [76], "not": [77], "been": [78], "comprehensively": [79], "studied.": [80], "In": [81], "this": [82], "work,": [83], "we": [84, 121, 161, 175], "systematically": [85], "investigate": [86], "Confidence-as-a-Reward": [87], "(CRew),": [88], "simple": [90], "yet": [91], "powerful": [92], "method": [94], "utilizes": [96], "token-level": [97], "in": [99], "model's": [101, 198], "final": [102], "answers": [103], "proxy": [106], "for": [107, 111], "reward,": [108], "especially": [109], "suitable": [110], "close-ended": [112], "tasks.": [113], "Through": [114], "experiments": [116], "on": [117, 130], "mathematical": [118], "tasks,": [120], "demonstrate": [122], "CRew": [124, 150, 164], "outperforms": [125, 203], "existing": [126, 204], "MATH500": [132], "RewardMATH": [134], "benchmarks,": [135], "even": [137], "surpasses": [138], "most": [139], "trained": [140], "models.": [142], "We": [143], "further": [144, 195], "identify": [145], "strong": [147], "correlation": [148], "scores": [151, 187], "actual": [154], "performance": [156], "model.": [159], "Additionally,": [160], "find": [162], "filter": [167], "high-quality": [168], "training": [169, 179], "data.": [170], "Building": [171], "upon": [172], "insights,": [174], "propose": [176], "CRew-DPO,": [177], "strategy": [180], "constructs": [182], "preference": [183], "from": [185], "combined": [188], "with": [189, 193], "correctness": [190], "signals.": [191], "Finetuning": [192], "CRew-DPO": [194], "enhances": [196], "judging": [199], "consistently": [202], "self-training": [205], "methods.": [206]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5737000107765198, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5737000107765198, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.1137000024318695, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12072", "display_name": "Machine Learning and Algorithms", "score": 0.03750000149011612, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.13501", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.13501", "pdf_url": "https://arxiv.org/pdf/2510.13501", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414743743", "doi": "https://doi.org/10.48550/arxiv.2509.16241", "title": "REAMS: Reasoning Enhanced Algorithm for Maths Solving", "publication_year": 2025, "publication_date": "2025-09-16", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Singh, Eishkaran", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Singh, Eishkaran", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5111218051", "display_name": "Tanav Singh Bajaj", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Bajaj, Tanav Singh", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": null, "display_name": "Nayak, Siddharth", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Nayak, Siddharth", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"The": [0], "challenges": [1, 134], "of": [2, 30, 98, 108, 125, 138], "solving": [3], "complex": [4, 141], "university-level": [5], "mathematics": [6], "problems,": [7], "particularly": [8], "those": [9], "from": [10, 19], "MIT,": [11], "and": [12, 16, 61, 68, 110, 131, 144], "Columbia": [13], "University": [14], "courses,": [15], "selected": [17], "tasks": [18], "the": [20, 28, 43, 105, 122, 133, 139], "MATH": [21], "dataset,": [22], "remain": [23], "a": [24, 54, 101, 112], "significant": [25, 123], "obstacle": [26], "in": [27, 39, 115], "field": [29], "artificial": [31], "intelligence.": [32], "Conventional": [33], "methods": [34], "have": [35], "consistently": [36], "fallen": [37], "short": [38], "this": [40, 50], "domain,": [41], "highlighting": [42], "need": [44], "for": [45, 71], "more": [46], "advanced": [47, 73, 126], "approaches.": [48], "In": [49], "paper,": [51], "we": [52], "introduce": [53], "language-based": [55], "solution": [56], "that": [57], "leverages": [58], "zero-shot": [59], "learning": [60], "mathematical": [62, 117, 142], "reasoning": [63], "to": [64, 129], "effectively": [65], "solve,": [66], "explain,": [67], "generate": [69], "solutions": [70], "these": [72], "math": [74], "problems.": [75], "By": [76], "integrating": [77], "program": [78], "synthesis,": [79], "our": [80], "method": [81], "reduces": [82], "reliance": [83], "on": [84], "large-scale": [85], "training": [86], "data": [87], "while": [88], "significantly": [89], "improving": [90], "problem-solving": [91], "accuracy.": [92], "Our": [93], "approach": [94], "achieves": [95], "an": [96], "accuracy": [97], "90.15%,": [99], "representing": [100], "substantial": [102], "improvement": [103], "over": [104], "previous": [106], "benchmark": [107], "81%": [109], "setting": [111], "new": [113], "standard": [114], "automated": [116], "problem-solving.": [118], "These": [119], "findings": [120], "highlight": [121], "potential": [124], "AI": [127], "methodologies": [128], "address": [130], "overcome": [132], "presented": [135], "by": [136], "some": [137], "most": [140], "courses": [143], "datasets.": [145]}, "primary_topic": {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.420199990272522, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.420199990272522, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10906", "display_name": "AI-based Problem Solving and Planning", "score": 0.11299999803304672, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13806", "display_name": "Mathematical and Computational Methods", "score": 0.09309999644756317, "subfield": {"id": "https://openalex.org/subfields/2611", "display_name": "Modeling and Simulation"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2509.16241", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2509.16241", "pdf_url": "https://arxiv.org/pdf/2509.16241", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414979462", "doi": "https://doi.org/10.48550/arxiv.2510.05825", "title": "Mitigating Premature Exploitation in Particle-based Monte Carlo for Inference-Time Scaling", "publication_year": 2025, "publication_date": "2025-10-07", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5077553088", "display_name": "Giorgio Giannone", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Giannone, Giorgio", "raw_affiliation_strings": ["Red Hat AI Innovation"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Red Hat AI Innovation", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5001313457", "display_name": "Guangxuan Xu", "orcid": "https://orcid.org/0000-0002-8784-7428", "observed_orcids": ["https://orcid.org/0000-0002-8784-7428"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Guangxuan", "raw_affiliation_strings": ["Red Hat AI Innovation"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Red Hat AI Innovation", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5115560574", "display_name": "Nikhil Shivakumar Nayak", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Nayak, Nikhil Shivakumar", "raw_affiliation_strings": ["Red Hat AI Innovation"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Red Hat AI Innovation", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5119915032", "display_name": "Rohan Mahesh Awhad", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Awhad, Rohan Mahesh", "raw_affiliation_strings": ["Red Hat AI Innovation"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Red Hat AI Innovation", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5033862118", "display_name": "Shivchander Sudalairaj", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Sudalairaj, Shivchander", "raw_affiliation_strings": ["Red Hat AI Innovation"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Red Hat AI Innovation", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5067011395", "display_name": "Kai Xu", "orcid": "https://orcid.org/0000-0003-0191-0375", "observed_orcids": ["https://orcid.org/0000-0003-0191-0375"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Kai", "raw_affiliation_strings": ["Red Hat AI Innovation"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Red Hat AI Innovation", "institution_ids": []}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5022716943", "display_name": "Akash Srivastava", "orcid": "https://orcid.org/0000-0002-5959-3926", "observed_orcids": ["https://orcid.org/0000-0002-5959-3926"]}, "institutions": [{"id": "https://openalex.org/I1341412227", "display_name": "IBM (United States)", "ror": "https://ror.org/05hh8d621", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I1341412227"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Srivastava, Akash", "raw_affiliation_strings": ["Core AI , IBM"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Core AI , IBM", "institution_ids": ["https://openalex.org/I1341412227"]}]}], "abstract_inverted_index": {"Inference-Time": [0], "Scaling": [1], "(ITS)": [2], "improves": [3], "language": [4], "models": [5], "by": [6, 34, 148, 159, 216], "allocating": [7], "more": [8], "computation": [9], "at": [10], "generation": [11], "time.": [12], "Particle": [13, 124], "Filtering": [14, 125], "(PF)": [15], "has": [16], "emerged": [17], "as": [18, 76], "a": [19, 98, 118, 177, 182, 203], "strong": [20, 197], "ITS": [21], "method": [22], "for": [23], "complex": [24], "mathematical": [25], "reasoning": [26, 46, 119], "tasks,": [27], "but": [28], "it": [29, 56, 157], "is": [30, 79], "vulnerable": [31], "when": [32, 154], "guided": [33], "process": [35], "reward": [36], "models,": [37], "which": [38], "often": [39], "assign": [40], "overconfident": [41, 108], "scores": [42], "early": [43], "in": [44, 102, 207], "the": [45, 91, 103, 115, 162, 218, 225], "process.": [47], "This": [48, 72], "causes": [49], "PF": [50], "to": [51, 59, 69, 107, 113, 134, 165, 180, 202, 232], "suffer": [52], "from": [53], "premature": [54], "exploitation:": [55], "myopically": [57], "commits": [58], "locally": [60], "promising": [61], "trajectories,": [62], "prunes": [63], "potentially": [64], "correct": [65], "hypotheses,": [66], "and": [67, 93, 110, 199], "converges": [68], "suboptimal": [70], "solutions.": [71, 234], "failure": [73], "mode,": [74], "known": [75], "particle": [77, 104, 146], "impoverishment,": [78], "especially": [80], "severe": [81], "under": [82], "constrained": [83], "computational": [84], "budgets.": [85], "To": [86], "address": [87], "this,": [88], "we": [89], "analyze": [90], "problem": [92], "identify": [94], "two": [95, 131], "root": [96], "causes:": [97], "lack": [99], "of": [100, 117, 220, 227], "diversity": [101, 151, 155], "set": [105], "due": [106], "resampling": [109, 163], "consequent": [111], "inability": [112], "assess": [114], "potential": [116, 184], "path.": [120], "We": [121], "introduce": [122], "Entropic": [123, 141], "(ePF),": [126], "an": [127, 170], "algorithm": [128], "that": [129], "integrates": [130], "new": [132], "techniques": [133], "solve": [135], "these": [136, 211], "issues.": [137], "The": [138, 168], "first": [139], "technique,": [140], "Annealing": [142], "(EA),": [143], "directly": [144], "mitigates": [145], "impoverishment": [147], "monitoring": [149], "search": [150], "via": [152], "entropy;": [153], "drops,": [156], "intervenes": [158], "dynamically": [160], "annealing": [161], "distribution": [164], "preserve": [166], "exploration.": [167], "second,": [169], "enhancement": [171], "called": [172], "Look-ahead": [173], "Modulation": [174], "(LaM),": [175], "adds": [176], "predictive": [178], "guide": [179], "evaluate": [181], "state's": [183], "based": [185], "on": [186], "its": [187], "successors.": [188], "On": [189], "several": [190], "challenging": [191], "math": [192], "benchmarks,": [193], "ePF": [194], "significantly": [195], "outperforms": [196], "baselines": [198], "achieves": [200], "up": [201], "50%": [204], "relative": [205], "improvement": [206], "task": [208], "reward.": [209], "Together,": [210], "methods": [212], "improve": [213], "PF's": [214], "resilience": [215], "balancing": [217], "exploration": [219], "diverse": [221], "solution": [222], "spaces": [223], "with": [224], "exploitation": [226], "high-reward": [228], "regions,": [229], "ultimately": [230], "leading": [231], "higher-quality": [233]}, "primary_topic": {"id": "https://openalex.org/T12056", "display_name": "Markov Chains and Monte Carlo Methods", "score": 0.3052000105381012, "subfield": {"id": "https://openalex.org/subfields/2613", "display_name": "Statistics and Probability"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T12056", "display_name": "Markov Chains and Monte Carlo Methods", "score": 0.3052000105381012, "subfield": {"id": "https://openalex.org/subfields/2613", "display_name": "Statistics and Probability"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.21199999749660492, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.09120000153779984, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.05825", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.05825", "pdf_url": "https://arxiv.org/pdf/2510.05825", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7139960534", "doi": "https://doi.org/10.34657/32244", "title": "MATRIX-MFO Tandem Workshop: Machine Learning and AI for Mathematics", "publication_year": 2025, "publication_date": "2025-01-01", "type": "article", "language": "en", "authorships": [], "abstract_inverted_index": {"This": [0], "workshop": [1], "explored": [2], "how": [3], "modern": [4], "machine": [5, 50], "learning": [6, 51], "can": [7], "both": [8], "accelerate": [9], "mathematical": [10, 58, 63], "discovery": [11], "and": [12, 38, 60, 69, 111, 118], "preserve": [13], "rigorous": [14], "standards.": [15], "It": [16], "focused": [17], "on": [18, 29], "three": [19], "angles:": [20], "using": [21, 32, 39, 49], "AI": [22, 36], "techniques": [23], "to": [24, 34, 66, 91], "help": [25], "mathematicians": [26], "make": [27], "advances": [28], "challenging": [30], "problems;": [31], "mathematics": [33], "understand": [35], "predictions;": [37], "deep-learning": [40], "models": [41], "for": [42, 55, 107, 115], "automated": [43, 96], "theorem": [44, 97, 104], "proving.": [45], "Key": [46], "discussions": [47], "included": [48], "as": [52], "a": [53], "tool": [54], "constructing": [56], "interesting": [57], "constructions": [59], "navigating": [61], "in": [62], "search": [64], "spaces,": [65], "uncover": [67], "conjectures": [68], "high-quality": [70, 108], "examples": [71], "(e.g.,": [72, 89, 101], "sphere": [73], "packings": [74], "via": [75, 79], "DiffuseBoost,": [76], "combinatorial": [77], "objects": [78], "AlphaEvolve);": [80], "Integrating": [81], "Large": [82], "Language": [83], "Models": [84], "(LLMs)": [85], "with": [86], "formal": [87], "systems": [88], "Lean/mathlib)": [90], "create": [92], "scalable,": [93], "certifiable": [94], "AI-based": [95], "prover;": [98], "Collaborative": [99], "formalization": [100], "the": [102], "Carleson": [103], "project),": [105], "autoformalization": [106], "supervised": [109], "data,": [110], "reinforcement": [112], "learning/search": [113], "methods": [114], "proof": [116], "generation": [117], "algorithmic": [119], "reasoning.": [120]}, "primary_topic": {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.5903000235557556, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.5903000235557556, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.21439999341964722, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.04470000043511391, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:oa.tib.eu:123456789/33176", "is_oa": true, "landing_page_url": "https://oa.tib.eu/renate/handle/123456789/33176", "pdf_url": null, "source": {"id": "https://openalex.org/S7407052981", "display_name": "RENATE", "issn_l": null, "issn": null, "is_oa": false, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I2802635041", "host_organization_name": "Technische Informationsbibliothek (TIB)", "host_organization_lineage": ["https://openalex.org/I2802635041"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by-nc-nd", "license_id": "https://openalex.org/licenses/cc-by-nc-nd", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": "Oberwolfach reports : OWR", "raw_type": "doc-type:Article"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415057821", "doi": "https://doi.org/10.48550/arxiv.2509.06948", "title": "Beyond Two-Stage Training: Cooperative SFT and RL for LLM Reasoning", "publication_year": 2025, "publication_date": "2025-09-08", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5101442863", "display_name": "Chen Liang", "orcid": "https://orcid.org/0009-0002-7454-9437", "observed_orcids": ["https://orcid.org/0009-0002-7454-9437"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Liang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Han, Xueting", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Han, Xueting", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100768717", "display_name": "Li Shen", "orcid": "https://orcid.org/0000-0002-0436-0199", "observed_orcids": ["https://orcid.org/0000-0002-0436-0199"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Shen, Li", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5052386069", "display_name": "Jing Bai", "orcid": "https://orcid.org/0009-0003-9650-3023", "observed_orcids": ["https://orcid.org/0009-0003-9650-3023"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Bai, Jing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5008208316", "display_name": "Kam\u2010Fai Wong", "orcid": "https://orcid.org/0000-0002-9427-5659", "observed_orcids": ["https://orcid.org/0000-0002-9427-5659"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wong, Kam-Fai", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Supervised": [0], "fine-tuning": [1], "(SFT)": [2], "and": [3, 32, 100, 103, 147, 159, 173, 178], "reinforcement": [4], "learning": [5], "with": [6], "verifiable": [7], "rewards": [8], "(RLVR)": [9], "are": [10, 53], "two": [11, 87, 115], "widely": [12], "used": [13], "post-training": [14], "paradigms": [15], "for": [16, 57], "improving": [17], "the": [18, 98, 114, 124], "reasoning": [19, 137, 172], "ability": [20], "of": [21, 126], "large": [22], "language": [23], "models": [24], "(LLMs).": [25], "Recent": [26], "methods": [27], "attempt": [28], "to": [29, 73, 106, 170, 176], "integrate": [30], "SFT": [31, 71, 99], "RLVR": [33], "in": [34, 69], "a": [35, 66, 93, 107, 119], "single": [36], "stage": [37], "by": [38, 76, 117], "reweighting": [39], "or": [40], "scheduling": [41], "their": [42], "objectives.": [43], "However,": [44], "such": [45], "coupling": [46], "can": [47], "be": [48], "counterproductive": [49], "because": [50], "supervised": [51], "updates": [52, 88], "not": [54], "uniformly": [55], "beneficial": [56], "reward": [58, 82, 125], "optimization.": [59, 83], "To": [60], "address": [61], "this,": [62], "we": [63], "propose": [64], "BRIDGE,": [65], "scalable": [67], "framework": [68], "which": [70], "learns": [72], "supervise": [74], "RL": [75, 101], "selectively": [77], "transferring": [78], "knowledge": [79], "that": [80, 96, 112, 167], "improves": [81], "Specifically,": [84], "BRIDGE": [85, 139, 168], "alternates": [86], "at": [89], "each": [90], "meta-training": [91], "step:": [92], "base-model": [94], "update": [95, 105], "fuses": [97], "gradients,": [102], "an": [104, 131], "lightweight": [108], "low-rank": [109], "adapter": [110], "(LoRA)": [111], "coordinates": [113], "objectives": [116], "maximizing": [118], "cooperative-gain": [120], "signal,": [121], "defined": [122], "as": [123], "joint": [127], "SFT-RL": [128], "training": [129, 162], "over": [130, 153], "RL-only": [132], "baseline.": [133], "Across": [134], "five": [135], "mathematical": [136], "benchmarks,": [138], "consistently": [140], "outperforms": [141], "two-stage": [142], "cold": [143], "start,": [144], "naive": [145], "mixing,": [146], "representative": [148], "single-stage": [149], "integration": [150], "baselines,": [151], "yielding": [152], "three": [154], "points": [155], "average": [156], "absolute": [157], "improvement": [158], "more": [160], "stable": [161], "dynamics.": [163], "We": [164], "further": [165], "show": [166], "extends": [169], "logical": [171], "generalizes": [174], "out-of-distribution": [175], "code": [177], "science": [179], "without": [180], "additional": [181], "training,": [182], "while": [183], "staying": [184], "robust": [185], "under": [186], "noisy": [187], "rewards.": [188]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5590999722480774, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5590999722480774, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.17100000381469727, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.05550000071525574, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2509.06948", "is_oa": true, "landing_page_url": "https://arxiv.org/abs/2509.06948", "pdf_url": "https://arxiv.org/pdf/2509.06948", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415035964", "doi": "https://doi.org/10.48550/arxiv.2505.20561", "title": "Beyond Markovian: Reflective Exploration via Bayes-Adaptive RL for LLM Reasoning", "publication_year": 2025, "publication_date": "2025-05-26", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5024816616", "display_name": "Shenao Zhang", "orcid": "https://orcid.org/0009-0006-5083-606X", "observed_orcids": ["https://orcid.org/0009-0006-5083-606X", "https://orcid.org/0009-0004-5233-3373"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Shenao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5006583739", "display_name": "Yaqing Wang", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Yaqing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5109012562", "display_name": "Yinxiao Liu", "orcid": "https://orcid.org/0009-0000-3181-2296", "observed_orcids": ["https://orcid.org/0009-0000-3181-2296"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Yinxiao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100618133", "display_name": "Tianqi Liu", "orcid": "https://orcid.org/0009-0003-5275-3741", "observed_orcids": ["https://orcid.org/0009-0003-5275-3741", "https://orcid.org/0009-0005-4758-8432"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Tianqi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Grabowski, Peter", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Grabowski, Peter", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5013936779", "display_name": "Eugene Ie", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ie, Eugene", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101934110", "display_name": "Zhaoran Wang", "orcid": "https://orcid.org/0000-0002-1824-2580", "observed_orcids": ["https://orcid.org/0000-0002-1824-2580"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Zhaoran", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5107646464", "display_name": "Yunxuan Li", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Yunxuan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "trained": [4], "via": [5], "Reinforcement": [6], "Learning": [7], "(RL)": [8], "have": [9], "exhibited": [10], "strong": [11], "reasoning": [12, 93, 187], "capabilities": [13], "and": [14, 21, 58, 147, 160, 173, 185, 200], "emergent": [15], "reflective": [16, 44, 92, 108], "behaviors,": [17], "such": [18], "as": [19, 24], "rethinking": [20], "error": [22], "correction,": [23], "a": [25, 84, 111, 121], "form": [26], "of": [27], "in-context": [28], "exploration.": [29], "However,": [30], "the": [31, 48, 52, 56, 80, 117, 130, 156, 165, 175], "Markovian": [32], "policy": [33, 49, 82], "obtained": [34], "from": [35], "conventional": [36, 193], "RL": [37, 71, 113, 194], "training": [38, 77, 131], "does": [39], "not": [40], "give": [41], "rise": [42], "to": [43, 63, 78, 158], "exploration": [45, 72, 109], "behaviors": [46], "since": [47], "depends": [50], "on": [51, 164, 171, 182], "history": [53], "only": [54, 74], "through": [55, 140], "state": [57], "therefore": [59], "has": [60], "no": [61], "incentive": [62], "enrich": [64], "identical": [65], "states": [66], "with": [67], "additional": [68], "context.": [69], "Instead,": [70], "is": [73, 101, 205], "useful": [75], "during": [76, 96], "learn": [79], "optimal": [81], "in": [83], "trial-and-error": [85], "manner.": [86], "Therefore,": [87], "it": [88, 100], "remains": [89], "unclear": [90], "whether": [91], "will": [94], "emerge": [95], "RL,": [97], "or": [98], "why": [99], "beneficial.": [102], "To": [103], "remedy": [104], "this,": [105], "we": [106], "recast": [107], "within": [110], "Bayesian": [112, 134], "framework,": [114], "which": [115], "optimizes": [116], "expected": [118], "return": [119], "under": [120], "posterior": [122], "distribution": [123], "over": [124], "Markov": [125], "decision": [126], "processes": [127], "induced": [128], "by": [129], "data.": [132], "This": [133], "formulation": [135], "admits": [136], "uncertainty-adaptive": [137], "policies": [138], "that,": [139], "belief": [141], "updates,": [142], "naturally": [143], "incentivize": [144], "information-gathering": [145], "actions": [146], "induce": [148], "self-reflection": [149], "behaviors.": [150], "Our": [151, 203], "resulting": [152], "algorithm,": [153], "BARL,": [154], "instructs": [155], "LLM": [157], "stitch": [159], "switch": [161], "strategies": [162], "based": [163], "observed": [166], "outcomes,": [167], "offering": [168], "principled": [169], "guidance": [170], "when": [172], "how": [174], "model": [176], "should": [177], "reflectively": [178], "explore.": [179], "Empirical": [180], "results": [181], "both": [183], "synthetic": [184], "mathematical": [186], "tasks": [188], "demonstrate": [189], "that": [190], "BARL": [191], "outperforms": [192], "approaches,": [195], "achieving": [196], "superior": [197], "test-time": [198], "performance": [199], "token": [201], "efficiency.": [202], "code": [204], "available": [206], "at": [207], "https://github.com/shenao-zhang/BARL.": [208]}, "primary_topic": {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.7324000000953674, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.7324000000953674, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.05999999865889549, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.05429999902844429, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2505.20561", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2505.20561", "pdf_url": "https://arxiv.org/pdf/2505.20561", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4416766928", "doi": "https://doi.org/10.48550/arxiv.2511.19942", "title": "Differential Smoothing Mitigates Sharpening and Improves LLM Reasoning", "publication_year": 2025, "publication_date": "2025-11-25", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5101293269", "display_name": "Jingchu Gai", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Gai, Jingchu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5111068871", "display_name": "Guanning Zeng", "orcid": "https://orcid.org/0009-0009-3783-9276", "observed_orcids": ["https://orcid.org/0009-0009-3783-9276"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zeng, Guanning", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5042509817", "display_name": "Huaqing Zhang", "orcid": "https://orcid.org/0000-0001-9180-7309", "observed_orcids": ["https://orcid.org/0000-0001-9180-7309"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Huaqing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5031731960", "display_name": "Aditi Raghunathan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Raghunathan, Aditi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"It": [0], "is": [1, 162], "widely": [2, 141], "recognized": [3], "that": [4, 97, 128, 159], "reinforcement": [5, 89], "learning": [6], "(RL)": [7], "fine-tuning": [8, 81], "of": [9, 28, 78], "large": [10], "language": [11], "models": [12, 168], "often": [13], "leads": [14], "to": [15, 30, 101, 107, 171, 194], "\\textit{diversity": [16], "collapse},": [17], "where": [18], "outputs": [19], "lack": [20], "variety.": [21], "Prior": [22], "work": [23], "has": [24], "proposed": [25], "a": [26, 69, 75, 86, 94, 121], "range": [27], "heuristics": [29, 151], "counteract": [31], "this": [32, 62, 117], "effect,": [33], "but": [34], "these": [35, 66], "methods": [36], "are": [37], "ad": [38], "hoc:": [39], "they": [40, 56, 155], "frequently": [41], "trade": [42], "off": [43], "correctness": [44, 132], "for": [45], "diversity,": [46, 134], "their": [47], "effectiveness": [48], "varies": [49], "across": [50, 174], "tasks,": [51], "and": [52, 88, 133, 153, 178, 190], "in": [53], "some": [54], "cases": [55], "even": [57], "contradict": [58], "one": [59], "another.": [60], "In": [61], "work,": [63], "we": [64, 92, 119], "place": [65], "observations": [67], "on": [68, 110, 116, 197], "rigorous": [70], "foundation.": [71], "We": [72], "first": [73], "provide": [74], "formal": [76], "proof": [77], "why": [79, 154], "RL": [80, 137], "exhibits": [82], "diversity": [83, 103], "collapse": [84, 104], "via": [85], "selection": [87], "bias.": [90], "Next,": [91], "make": [93], "key": [95], "observation": [96], "any": [98], "reward": [99], "modification": [100], "address": [102], "only": [105], "needs": [106], "be": [108], "applied": [109], "the": [111], "correct": [112], "trajectories.": [113], "Building": [114], "directly": [115], "analysis,": [118], "introduce": [120], "principled": [122], "method": [123], "--": [124, 127], "\\textit{differential": [125], "smoothing}": [126], "provably": [129], "improves": [130, 187], "both": [131, 188], "outperforming": [135], "vanilla": [136], "as": [138, 140], "well": [139], "used": [142], "entropy-based": [143], "heuristics.": [144], "Our": [145], "theory": [146], "precisely": [147], "characterizes": [148], "when": [149], "existing": [150], "help": [152], "fail,": [156], "while": [157], "showing": [158], "differential": [160], "smoothing": [161, 186], "universally": [163], "superior.": [164], "Extensive": [165], "experiments": [166], "with": [167, 192], "from": [169], "1B": [170], "7B": [172], "parameters,": [173], "domains": [175], "including": [176], "CountDown": [177], "real-world": [179], "mathematical": [180], "reasoning,": [181], "demonstrate": [182], "consistent": [183], "gains.": [184], "Differential": [185], "Pass@1": [189], "Pass@k,": [191], "up": [193], "6.7\\%": [195], "improvements": [196], "AIME24": [198], "dataset.": [199]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.48539999127388, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.48539999127388, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.23559999465942383, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.0803999975323677, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2511.19942", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2511.19942", "pdf_url": "https://arxiv.org/pdf/2511.19942", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414973451", "doi": "https://doi.org/10.48550/arxiv.2504.07086", "title": "A Sober Look at Progress in Language Model Reasoning: Pitfalls and Paths to Reproducibility", "publication_year": 2025, "publication_date": "2025-04-09", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5107077701", "display_name": "Andreas Hochlehnert", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4401726832", "display_name": "T\u00fcbingen AI Center", "ror": "https://ror.org/0107nyd78", "country_code": null, "type": "education", "lineage": ["https://openalex.org/I149899117", "https://openalex.org/I4210135521", "https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}, {"id": "https://openalex.org/I8087733", "display_name": "University of T\u00fcbingen", "ror": "https://ror.org/03a1kwz48", "country_code": "DE", "type": "education", "lineage": ["https://openalex.org/I8087733"]}], "countries": ["DE"], "is_corresponding": false, "raw_author_name": "Hochlehnert, Andreas", "raw_affiliation_strings": ["T\u00fcbingen AI Center , University of T\u00fcbingen"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "T\u00fcbingen AI Center , University of T\u00fcbingen", "institution_ids": ["https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5114438807", "display_name": "Hardik Bhatnagar", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4401726832", "display_name": "T\u00fcbingen AI Center", "ror": "https://ror.org/0107nyd78", "country_code": null, "type": "education", "lineage": ["https://openalex.org/I149899117", "https://openalex.org/I4210135521", "https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}, {"id": "https://openalex.org/I8087733", "display_name": "University of T\u00fcbingen", "ror": "https://ror.org/03a1kwz48", "country_code": "DE", "type": "education", "lineage": ["https://openalex.org/I8087733"]}], "countries": ["DE"], "is_corresponding": false, "raw_author_name": "Bhatnagar, Hardik", "raw_affiliation_strings": ["T\u00fcbingen AI Center , University of T\u00fcbingen"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "T\u00fcbingen AI Center , University of T\u00fcbingen", "institution_ids": ["https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5108719787", "display_name": "Vishaal Udandarao", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4401726832", "display_name": "T\u00fcbingen AI Center", "ror": "https://ror.org/0107nyd78", "country_code": null, "type": "education", "lineage": ["https://openalex.org/I149899117", "https://openalex.org/I4210135521", "https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}, {"id": "https://openalex.org/I8087733", "display_name": "University of T\u00fcbingen", "ror": "https://ror.org/03a1kwz48", "country_code": "DE", "type": "education", "lineage": ["https://openalex.org/I8087733"]}], "countries": ["DE"], "is_corresponding": false, "raw_author_name": "Udandarao, Vishaal", "raw_affiliation_strings": ["T\u00fcbingen AI Center , University of T\u00fcbingen", "University of Cambridge Leaderboard Code Eval"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "T\u00fcbingen AI Center , University of T\u00fcbingen", "institution_ids": ["https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}, {"raw_affiliation_string": "University of Cambridge Leaderboard Code Eval", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102018867", "display_name": "Samuel Albanie", "orcid": "https://orcid.org/0000-0003-1732-9198", "observed_orcids": ["https://orcid.org/0000-0003-1732-9198"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Albanie, Samuel", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5113274335", "display_name": "Ameya Prabhu", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4401726832", "display_name": "T\u00fcbingen AI Center", "ror": "https://ror.org/0107nyd78", "country_code": null, "type": "education", "lineage": ["https://openalex.org/I149899117", "https://openalex.org/I4210135521", "https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}, {"id": "https://openalex.org/I8087733", "display_name": "University of T\u00fcbingen", "ror": "https://ror.org/03a1kwz48", "country_code": "DE", "type": "education", "lineage": ["https://openalex.org/I8087733"]}], "countries": ["DE"], "is_corresponding": false, "raw_author_name": "Prabhu, Ameya", "raw_affiliation_strings": ["T\u00fcbingen AI Center , University of T\u00fcbingen"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "T\u00fcbingen AI Center , University of T\u00fcbingen", "institution_ids": ["https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5061457780", "display_name": "Matthias Bethge", "orcid": "https://orcid.org/0000-0002-6417-7812", "observed_orcids": ["https://orcid.org/0000-0002-6417-7812", "https://orcid.org/0009-0008-5601-6120"]}, "institutions": [{"id": "https://openalex.org/I4401726832", "display_name": "T\u00fcbingen AI Center", "ror": "https://ror.org/0107nyd78", "country_code": null, "type": "education", "lineage": ["https://openalex.org/I149899117", "https://openalex.org/I4210135521", "https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}, {"id": "https://openalex.org/I8087733", "display_name": "University of T\u00fcbingen", "ror": "https://ror.org/03a1kwz48", "country_code": "DE", "type": "education", "lineage": ["https://openalex.org/I8087733"]}], "countries": ["DE"], "is_corresponding": false, "raw_author_name": "Bethge, Matthias", "raw_affiliation_strings": ["T\u00fcbingen AI Center , University of T\u00fcbingen"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "T\u00fcbingen AI Center , University of T\u00fcbingen", "institution_ids": ["https://openalex.org/I4401726832", "https://openalex.org/I8087733"]}]}], "abstract_inverted_index": {"Reasoning": [0], "has": [1], "emerged": [2], "as": [3], "the": [4, 154], "next": [5], "major": [6], "frontier": [7], "for": [8, 169, 176], "language": [9], "models": [10], "(LMs),": [11], "with": [12, 28, 103], "rapid": [13], "advances": [14], "from": [15], "both": [16], "academic": [17], "and": [18, 51, 71, 74, 108, 118, 166], "industrial": [19], "labs.": [20], "However,": [21], "this": [22, 43, 112], "progress": [23], "often": [24], "outpaces": [25], "methodological": [26], "rigor,": [27], "many": [29], "evaluations": [30], "relying": [31], "on": [32, 85, 138], "benchmarking": [33], "practices": [34, 107], "that": [35, 53, 120], "lack": [36], "transparency,": [37], "robustness,": [38], "or": [39, 88], "statistical": [40], "grounding.": [41], "In": [42, 143], "work,": [44], "we": [45, 97, 114, 156, 161], "conduct": [46], "a": [47, 99], "comprehensive": [48], "empirical": [49], "study": [50], "find": [52, 119], "current": [54], "mathematical": [55], "reasoning": [56, 170], "benchmarks": [57, 140], "are": [58, 133], "highly": [59], "sensitive": [60], "to": [61, 135], "subtle": [62], "implementation": [63], "choices--including": [64], "decoding": [65], "parameters,": [66], "random": [67], "seeds,": [68], "prompt": [69], "formatting,": [70], "even": [72], "hardware": [73], "software": [75], "configurations.": [76], "Performance": [77], "gains": [78], "reported": [79], "in": [80, 153], "recent": [81, 116], "studies": [82], "frequently": [83], "hinge": [84], "unclear": [86], "comparisons": [87], "unreported": [89], "sources": [90], "of": [91], "variance.": [92], "To": [93, 158], "address": [94], "these": [95], "issues,": [96], "propose": [98], "standardized": [100], "evaluation": [101], "framework": [102], "clearly": [104], "defined": [105], "best": [106], "reporting": [109], "standards.": [110], "Using": [111], "framework,": [113], "reassess": [115], "methods": [117, 148], "most": [121], "reinforcement": [122], "learning": [123], "(RL)": [124], "approaches": [125], "yield": [126], "only": [127], "modest": [128], "improvements--far": [129], "below": [130], "prior": [131], "claims--and": [132], "prone": [134], "overfitting,": [136], "especially": [137], "small-scale": [139], "like": [141], "AIME'24.": [142], "contrast,": [144], "supervised": [145], "finetuning": [146], "(SFT)": [147], "show": [149], "consistently": [150], "stronger": [151], "generalization": [152], "settings": [155], "study.": [157], "foster": [159], "reproducibility,": [160], "release": [162], "all": [163], "code,": [164], "prompts,": [165], "model": [167], "outputs,": [168], "benchmarks,": [171], "establishing": [172], "more": [173], "rigorous": [174], "foundations": [175], "future": [177], "work.": [178]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7562999725341797, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7562999725341797, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.1054999977350235, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.013700000010430813, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2504.07086", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2504.07086", "pdf_url": "https://arxiv.org/pdf/2504.07086", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415097482", "doi": "https://doi.org/10.48550/arxiv.2510.06870", "title": "$\u03bb$-GRPO: Unifying the GRPO Frameworks with Learnable Token Preferences", "publication_year": 2025, "publication_date": "2025-10-08", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Wang, Yining", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Yining", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5036110657", "display_name": "Jinman Zhao", "orcid": "https://orcid.org/0009-0001-1511-8875", "observed_orcids": ["https://orcid.org/0009-0001-1511-8875"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhao, Jinman", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Zhao, Chuangxin", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhao, Chuangxin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Guan, Shuhao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Guan, Shuhao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5052428595", "display_name": "Gerald Penn", "orcid": "https://orcid.org/0000-0003-3553-8305", "observed_orcids": ["https://orcid.org/0000-0003-3553-8305"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Penn, Gerald", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5055004432", "display_name": "Shuang Liu", "orcid": "https://orcid.org/0009-0007-8540-2496", "observed_orcids": ["https://orcid.org/0009-0007-8540-2496"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Shinan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Reinforcement": [0, 22], "Learning": [1, 23], "with": [2, 24, 39, 185], "Human": [3], "Feedback": [4], "(RLHF)": [5], "has": [6, 28], "been": [7], "the": [8, 13, 34, 60, 78, 99, 103, 124, 128, 212, 220], "dominant": [9], "approach": [10], "for": [11], "improving": [12], "reasoning": [14, 180], "capabilities": [15], "of": [16, 69, 102, 126, 224], "Large": [17], "Language": [18], "Models": [19], "(LLMs).": [20], "Recently,": [21], "Verifiable": [25], "Rewards": [26], "(RLVR)": [27], "simplified": [29], "this": [30, 120], "paradigm": [31], "by": [32, 195], "replacing": [33], "reward": [35, 79], "and": [36, 83, 95, 110, 146, 164, 175, 188, 198, 222], "value": [37], "models": [38, 184], "rule-based": [40], "verifiers.": [41], "A": [42], "prominent": [43], "example": [44], "is": [45, 63], "Group": [46], "Relative": [47], "Policy": [48], "Optimization": [49], "(GRPO).": [50], "However,": [51], "GRPO": [52, 174], "inherently": [53], "suffers": [54], "from": [55], "a": [56, 70, 73, 143, 148], "length": [57], "bias,": [58], "since": [59], "same": [61], "advantage": [62], "uniformly": [64], "assigned": [65], "to": [66, 87, 130, 160, 201, 211], "all": [67], "tokens": [68, 82], "response.": [71], "As": [72], "result,": [74], "longer": [75], "responses": [76], "distribute": [77], "over": [80, 172], "more": [81], "thus": [84], "contribute": [85], "disproportionately": [86], "gradient": [88], "updates.": [89], "Several": [90], "variants,": [91], "such": [92], "as": [93], "DAPO": [94, 176], "Dr.": [96], "GRPO,": [97, 202], "modify": [98], "token-level": [100, 155], "aggregation": [101], "loss,": [104], "yet": [105], "these": [106, 205], "methods": [107], "remain": [108], "heuristic": [109], "offer": [111], "limited": [112], "interpretability": [113], "regarding": [114], "their": [115], "implicit": [116], "token": [117, 134, 226], "preferences.": [118, 227], "In": [119], "work,": [121], "we": [122, 165], "explore": [123], "possibility": [125], "allowing": [127], "model": [129], "learn": [131], "its": [132], "own": [133], "preference": [135], "during": [136], "optimization.": [137], "We": [138, 157], "unify": [139], "existing": [140], "frameworks": [141], "under": [142], "single": [144], "formulation": [145], "introduce": [147], "learnable": [149], "parameter": [150], "$\u03bb$": [151], "that": [152, 167], "adaptively": [153], "controls": [154], "weighting.": [156], "use": [158], "$\u03bb$-GRPO": [159, 168, 191], "denote": [161], "our": [162], "method,": [163], "find": [166], "achieves": [169], "consistent": [170], "improvements": [171], "vanilla": [173], "on": [177], "multiple": [178], "mathematical": [179], "benchmarks.": [181], "On": [182], "Qwen2.5": [183], "1.5B,": [186], "3B,": [187], "7B": [189], "parameters,": [190], "improves": [192], "average": [193], "accuracy": [194], "$+1.9\\%$,": [196], "$+1.0\\%$,": [197], "$+1.7\\%$": [199], "compared": [200], "respectively.": [203], "Importantly,": [204], "gains": [206], "come": [207], "without": [208], "any": [209], "modifications": [210], "training": [213], "data": [214], "or": [215], "additional": [216], "computational": [217], "cost,": [218], "highlighting": [219], "effectiveness": [221], "practicality": [223], "learning": [225]}, "primary_topic": {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.448199987411499, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.448199987411499, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.20059999823570251, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.10949999839067459, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.06870", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.06870", "pdf_url": "https://arxiv.org/pdf/2510.06870", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4416245178", "doi": "https://doi.org/10.48550/arxiv.2510.20272", "title": "Limits of PRM-Guided Tree Search for Mathematical Reasoning with LLMs", "publication_year": 2025, "publication_date": "2025-10-23", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5087483149", "display_name": "Tristan Cinquin", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Cinquin, Tristan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5065203625", "display_name": "Geoff Pleiss", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Pleiss, Geoff", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5078204924", "display_name": "Agustinus Kristiadi", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kristiadi, Agustinus", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"While": [0], "chain-of-thought": [1], "prompting": [2], "with": [3, 75, 124], "Best-of-N": [4], "(BoN)": [5], "selection": [6], "has": [7], "become": [8], "popular": [9], "for": [10], "mathematical": [11, 60, 71, 159], "reasoning": [12, 61, 125, 160], "in": [13, 161], "large": [14], "language": [15], "models": [16], "(LLMs),": [17], "its": [18, 76], "linear": [19], "structure": [20], "fails": [21], "to": [22, 40], "capture": [23], "the": [24, 48], "branching": [25], "and": [26, 52, 105, 120, 127], "exploratory": [27], "nature": [28], "of": [29, 133], "complex": [30], "problem-solving.": [31], "In": [32], "this": [33], "work,": [34], "we": [35, 83], "propose": [36], "an": [37], "adaptive": [38], "algorithm": [39], "maximize": [41], "process": [42], "reward": [43, 149], "model": [44], "(PRM)": [45], "scores": [46], "over": [47, 95], "intractable": [49], "action": [50], "space,": [51], "investigate": [53], "whether": [54], "PRM-guided": [55, 87, 110], "tree": [56, 88, 103, 111, 139, 154], "search": [57, 89, 104, 107, 112, 155], "can": [58, 156], "improve": [59], "by": [62], "exploring": [63], "multiple": [64], "partial": [65], "solution": [66], "paths.": [67], "Across": [68], "$23$": [69], "diverse": [70], "problems": [72], "using": [73], "Qwen2.5-Math-7B-Instruct": [74], "associated": [77], "PRM": [78, 145], "as": [79], "a": [80], "case": [81], "study,": [82], "find": [84], "that:": [85], "(1)": [86], "shows": [90], "no": [91], "statistically": [92], "significant": [93], "improvements": [94], "BoN": [96], "despite": [97], "higher": [98], "costs,": [99], "(2)": [100], "Monte": [101], "Carlo": [102], "beam": [106], "outperform": [108], "other": [109], "methods,": [113], "(3)": [114], "PRMs": [115, 129], "poorly": [116, 131], "approximate": [117], "state": [118], "values": [119], "their": [121], "reliability": [122], "degrades": [123], "depth,": [126], "(4)": [128], "generalize": [130], "out": [132], "distribution.": [134], "This": [135], "underperformance": [136], "stems": [137], "from": [138], "search's": [140], "greater": [141], "reliance": [142], "on": [143], "unreliable": [144], "scores,": [146], "suggesting": [147], "different": [148], "modeling": [150], "is": [151], "necessary": [152], "before": [153], "effectively": [157], "enhance": [158], "LLMs.": [162]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.3361999988555908, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.3361999988555908, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.12049999833106995, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10906", "display_name": "AI-based Problem Solving and Planning", "score": 0.1143999993801117, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.20272", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.20272", "pdf_url": "https://arxiv.org/pdf/2510.20272", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414814642", "doi": "https://doi.org/10.48550/arxiv.2510.01253", "title": "OR-Toolformer: Modeling and Solving Operations Research Problems with Tool Augmented Large Language Models", "publication_year": 2025, "publication_date": "2025-09-24", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5114218131", "display_name": "Jindong Zhang", "orcid": "https://orcid.org/0009-0002-5158-909X", "observed_orcids": ["https://orcid.org/0009-0002-5158-909X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Jianzhang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5023169037", "display_name": "Jialong Zhou", "orcid": "https://orcid.org/0009-0008-5476-3448", "observed_orcids": ["https://orcid.org/0009-0008-5476-3448"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhou, Jialong", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100326622", "display_name": "Chuang Liu", "orcid": "https://orcid.org/0000-0003-3745-1051", "observed_orcids": ["https://orcid.org/0000-0003-3745-1051", "https://orcid.org/0000-0002-2087-7673"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Chuang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Large": [0], "language": [1], "models": [2, 22], "(LLMs)": [3], "demonstrate": [4], "strong": [5], "mathematical": [6], "reasoning,": [7], "but": [8], "reliance": [9], "on": [10, 80], "closed-source": [11], "APIs": [12], "for": [13, 108], "OR": [14, 44, 83, 112], "tasks": [15], "raises": [16], "privacy": [17], "concerns,": [18], "and": [19, 47, 110, 115], "training": [20], "open-source": [21], "from": [23], "scratch": [24], "incurs": [25], "high": [26], "compute": [27], "costs.": [28], "We": [29], "introduce": [30], "OR-Toolformer,": [31], "which": [32], "fine-tunes": [33], "Llama-3.1-8B-Instruct": [34], "with": [35, 51], "a": [36, 91], "semi-automatic": [37], "data": [38], "synthesis": [39], "pipeline": [40], "that": [41], "generates": [42], "diverse": [43], "problem-answer": [45], "pairs": [46], "augments": [48], "the": [49, 96, 102], "model": [50], "external": [52], "solvers": [53], "to": [54, 67], "produce": [55], "API": [56], "calls.": [57], "On": [58], "three": [59], "of": [60, 104], "four": [61], "standard": [62], "benchmarks,": [63], "OR-Toolformer": [64], "achieves": [65], "up": [66], "80.1%": [68], "execution": [69], "accuracy,": [70, 90], "exceeding": [71], "size-matched": [72], "baselines": [73], "by": [74], "over": [75, 95], "4.3%.": [76], "In": [77], "zero-shot": [78], "evaluation": [79], "two": [81], "unseen": [82], "problem": [84, 113], "types,": [85], "it": [86], "attains": [87], "54%": [88], "average": [89], "21": [92], "percentage-point": [93], "improvement": [94], "strongest": [97], "baseline.": [98], "These": [99], "findings": [100], "validate": [101], "efficacy": [103], "tool-augmented": [105], "fine-tuning": [106], "LLMs": [107], "accurate": [109], "generalizable": [111], "modeling": [114], "solving.": [116]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7002000212669373, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7002000212669373, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.1266999989748001, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10906", "display_name": "AI-based Problem Solving and Planning", "score": 0.025699999183416367, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.01253", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.01253", "pdf_url": "https://arxiv.org/pdf/2510.01253", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414814585", "doi": "https://doi.org/10.48550/arxiv.2510.01227", "title": "EEFSUVA: A New Mathematical Olympiad Benchmark", "publication_year": 2025, "publication_date": "2025-09-23", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5119842918", "display_name": "Nicole N Khatibi", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Khatibi, Nicole N", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5119842919", "display_name": "Daniil A. Radamovich", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Radamovich, Daniil A.", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5104321680", "display_name": "Michael P. Brenner", "orcid": "https://orcid.org/0000-0002-5673-7947", "observed_orcids": ["https://orcid.org/0000-0002-5673-7947"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Brenner, Michael P.", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Recent": [0], "breakthroughs": [1], "have": [2], "spurred": [3], "claims": [4, 27], "that": [5, 142], "large": [6], "language": [7], "models": [8, 61], "(LLMs)": [9], "match": [10], "gold": [11], "medal": [12], "Olympiad": [13, 54], "to": [14, 34, 65, 118, 154], "graduate": [15], "level": [16], "proficiency": [17], "on": [18, 73, 151], "mathematics": [19], "benchmarks.": [20, 157], "In": [21], "this": [22], "work,": [23], "we": [24, 86], "examine": [25], "these": [26, 46], "in": [28, 136], "detail": [29], "and": [30, 56, 69, 97, 103, 121, 176], "assess": [31], "the": [32, 51, 104, 107, 119, 162], "extent": [33], "which": [35], "current": [36], "benchmarks": [37], "capture": [38], "genuine": [39], "LLM": [40], "mathematical": [41, 84, 174], "reasoning.": [42], "The": [43], "composition": [44], "of": [45, 83, 100, 115, 165, 173], "benchmarks,": [47], "primarily": [48], "drawing": [49], "from": [50, 93, 106], "International": [52], "Mathematics": [53], "(IMO)": [55], "related": [57], "competitions,": [58], "may": [59], "overstate": [60], "reasoning": [62, 175], "ability": [63], "due": [64], "potential": [66, 163], "data": [67], "contamination": [68], "a": [70, 79, 89, 147, 170], "narrow": [71], "focus": [72], "familiar": [74], "problem": [75], "types.": [76], "To": [77], "enable": [78], "more": [80], "holistic": [81], "assessment": [82, 172], "understanding,": [85], "introduce": [87], "EEFSUVA,": [88], "novel": [90], "benchmark": [91], "curated": [92], "under": [94], "circulated": [95], "regional": [96], "national": [98], "Olympiads": [99], "Eastern": [101], "Europe": [102], "countries": [105], "former": [108], "Soviet": [109], "Union.": [110], "These": [111, 158], "contests": [112], "feature": [113], "problems": [114, 131], "comparable": [116], "difficulty": [117], "IMO": [120], "are": [122, 132], "renowned": [123], "for": [124, 169, 177], "demanding": [125], "nonstandard": [126], "problem-solving": [127], "techniques,": [128], "yet": [129], "their": [130], "far": [133], "less": [134], "prevalent": [135], "online": [137], "corpora.": [138], "Preliminary": [139], "results": [140], "suggest": [141, 161], "even": [143], "state-of-the-art": [144], "LLMs": [145], "exhibit": [146], "notable": [148], "performance": [149], "decline": [150], "EEFSUVA": [152], "relative": [153], "other": [155], "Olympiad-style": [156], "findings": [159], "also": [160], "importance": [164], "broader": [166], "evaluation": [167], "datasets": [168], "fuller": [171], "guiding": [178], "future": [179], "model": [180], "development.": [181]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.4465000033378601, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.4465000033378601, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.19840000569820404, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.12620000541210175, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.01227", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.01227", "pdf_url": "https://arxiv.org/pdf/2510.01227", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4416936859", "doi": "https://doi.org/10.48550/arxiv.2512.00466", "title": "SCALE: Selective Resource Allocation for Overcoming Performance Bottlenecks in Mathematical Test-time Scaling", "publication_year": 2025, "publication_date": "2025-11-29", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5100740881", "display_name": "Yang Xiao", "orcid": "https://orcid.org/0000-0002-7739-4146", "observed_orcids": ["https://orcid.org/0000-0002-7739-4146", "https://orcid.org/0009-0005-1382-7386"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xiao, Yang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5002321609", "display_name": "Chunpu Xu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Chunpu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100566675", "display_name": "Ruifeng Yuan", "orcid": "https://orcid.org/0000-0002-5621-9648", "observed_orcids": ["https://orcid.org/0000-0002-5621-9648"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yuan, Ruifeng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Wang, Jiashuo", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Jiashuo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100408983", "display_name": "Wenjie Li", "orcid": "https://orcid.org/0000-0002-7360-8864", "observed_orcids": ["https://orcid.org/0000-0002-7360-8864", "https://orcid.org/0009-0005-7397-031X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Wenjie", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100355025", "display_name": "Pengfei Liu", "orcid": "https://orcid.org/0009-0008-6932-7091", "observed_orcids": ["https://orcid.org/0009-0008-6932-7091"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Pengfei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Test-time": [0], "compute": [1], "scaling": [2, 164, 192], "has": [3], "emerged": [4], "as": [5], "a": [6, 74, 187], "powerful": [7], "paradigm": [8], "for": [9, 120, 126], "enhancing": [10], "mathematical": [11], "reasoning": [12, 34, 95], "in": [13, 190], "large": [14], "language": [15], "models": [16], "(LLMs)": [17], "by": [18, 65, 184], "allocating": [19], "additional": [20, 58], "computational": [21, 59, 79, 182], "resources": [22, 60, 80, 138], "during": [23], "inference.": [24], "However,": [25], "current": [26, 198], "methods": [27], "employ": [28], "uniform": [29, 52, 163], "resource": [30, 154], "distribution": [31], "across": [32], "all": [33], "sub-problems,": [35, 96, 111], "creating": [36], "fundamental": [37, 195], "bottlenecks": [38, 56], "where": [39, 57], "challenging": [40, 110, 140], "sub-problems": [41, 122, 141], "receive": [42], "insufficient": [43], "attention": [44], "while": [45, 142, 180], "routine": [46, 106, 144], "operations": [47, 107, 145], "consume": [48], "disproportionate": [49], "resources.": [50], "This": [51], "allocation": [53], "creates": [54], "performance": [55, 150], "yield": [61], "diminishing": [62], "returns.": [63], "Inspired": [64], "dual-process": [66], "theory,": [67], "we": [68], "propose": [69], "\\textbf{SCALE}": [70], "(Selective": [71], "Resource": [72], "Allocation),": [73], "framework": [75], "that": [76, 159, 193], "selectively": [77], "allocates": [78], "based": [81], "on": [82, 139, 178], "sub-problem": [83, 102], "difficulty.": [84], "SCALE": [85, 147, 160], "operates": [86], "through": [87], "four": [88], "stages:": [89], "(1)": [90], "problem": [91], "decomposition": [92], "into": [93], "sequential": [94, 131], "(2)": [97], "difficulty": [98], "assessment": [99], "of": [100, 169, 197], "each": [101], "to": [103, 171, 176], "distinguish": [104], "between": [105, 117], "and": [108, 123, 129], "computationally": [109], "(3)": [112], "selective": [113], "processing": [114, 143], "mode": [115], "assignment": [116], "System": [118, 124], "1": [119], "simple": [121], "2": [125], "complex": [127], "ones,": [128], "(4)": [130], "execution": [132], "with": [133, 152], "context": [134], "propagation.": [135], "By": [136], "concentrating": [137], "efficiently,": [146], "achieves": [148], "substantial": [149], "improvements": [151, 168], "superior": [153], "utilization.": [155], "Extensive": [156], "experiments": [157], "demonstrate": [158], "significantly": [161], "outperforms": [162], "baselines,": [165], "achieving": [166], "accuracy": [167], "up": [170], "13.75": [172], "percentage": [173], "points": [174], "(57.50%": [175], "71.25%": [177], "AIME25)": [179], "reducing": [181], "costs": [183], "33%-53%,": [185], "representing": [186], "major": [188], "advance": [189], "test-time": [191], "addresses": [194], "limitations": [196], "approaches.": [199]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.3621000051498413, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.3621000051498413, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.28369998931884766, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.05849999934434891, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2512.00466", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2512.00466", "pdf_url": "https://arxiv.org/pdf/2512.00466", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7117627197", "doi": "https://doi.org/10.1109/tnnls.2025.3644134", "title": "Large Language Models Are Multitask Chain-of-Thought Prompting Optimizers", "publication_year": 2025, "publication_date": "2025-12-30", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5064142832", "display_name": "Feihu Jin", "orcid": "https://orcid.org/0009-0001-5551-7723", "observed_orcids": ["https://orcid.org/0009-0001-5551-7723"]}, "institutions": [{"id": "https://openalex.org/I20231570", "display_name": "Peking University", "ror": "https://ror.org/02v51f717", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I20231570"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Feihu Jin", "raw_affiliation_strings": ["School of Intelligence Science and Technology and the Institute for Artificial Intelligence, Peking University, Beijing, China"], "raw_orcid": "https://orcid.org/0009-0001-5551-7723", "affiliations": [{"raw_affiliation_string": "School of Intelligence Science and Technology and the Institute for Artificial Intelligence, Peking University, Beijing, China", "institution_ids": ["https://openalex.org/I20231570"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5121524429", "display_name": "Ying Tan", "orcid": "https://orcid.org/0000-0001-8243-4731", "observed_orcids": ["https://orcid.org/0000-0001-8243-4731"]}, "institutions": [{"id": "https://openalex.org/I20231570", "display_name": "Peking University", "ror": "https://ror.org/02v51f717", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I20231570"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Ying Tan", "raw_affiliation_strings": ["School of Intelligence Science and Technology and the Institute for Artificial Intelligence, Peking University, Beijing, China"], "raw_orcid": "https://orcid.org/0000-0001-8243-4731", "affiliations": [{"raw_affiliation_string": "School of Intelligence Science and Technology and the Institute for Artificial Intelligence, Peking University, Beijing, China", "institution_ids": ["https://openalex.org/I20231570"]}]}], "abstract_inverted_index": {"Large": [0], "language": [1, 115], "models": [2], "(LLMs)": [3], "have": [4, 61], "achieved": [5], "striking": [6], "performance": [7], "across": [8], "a": [9, 43, 84, 93, 104], "broad": [10], "range": [11], "of": [12, 18, 45, 111, 155], "reasoning": [13, 161, 167], "benchmarks,": [14], "yet": [15, 95], "the": [16, 77, 100, 133], "quality": [17], "their": [19], "outputs": [20], "remains": [21], "acutely": [22], "sensitive": [23], "to": [24, 73, 89, 157], "prompt": [25, 30, 51, 59], "design.": [26], "A": [27], "meticulously": [28], "engineered": [29], "can": [31], "coax": [32], "an": [33], "LLM": [34, 101], "into": [35, 47], "correctly": [36], "answering": [37], "even": [38], "highly": [39], "complex": [40], "questions,": [41], "prompting": [42, 56], "surge": [44], "research": [46], "techniques": [48], "that": [49, 124, 146], "boost": [50], "efficacy.": [52], "Manual": [53], "chain-of-thought": [54], "(CoT)": [55], "and": [57, 118, 163], "automated": [58], "generation": [60], "emerged": [62], "as": [63, 103], "leading": [64], "strategies.": [65], "However,": [66], "CoT": [67], "exemplars": [68], "must": [69], "be": [70], "painstakingly": [71], "tailored": [72], "each": [74], "task,": [75], "making": [76], "process": [78], "labor-intensive,": [79], "while": [80], "prompts": [81, 112, 126, 135], "optimized": [82], "for": [83], "single": [85], "task": [86, 116], "often": [87], "fail": [88], "generalize.": [90], "We": [91], "introduce": [92], "simple": [94], "powerful": [96], "alternative:": [97], "by": [98], "treating": [99], "itself": [102], "multitask": [105], "optimizer,": [106], "we": [107, 130], "enable": [108], "iterative": [109], "self-refinement": [110], "through": [113], "natural": [114], "descriptions": [117], "few-shot": [119], "in-context": [120], "learning": [121], "(ICL).": [122], "Recognizing": [123], "individual": [125], "exhibit": [127], "task-dependent": [128], "sensitivity,": [129], "further": [131], "ensemble": [132], "top-performing": [134], "at": [136], "inference": [137], "time.": [138], "Empirical": [139], "evaluation": [140], "with": [141], "several": [142], "state-of-the-art": [143], "LLMs": [144], "shows": [145], "our": [147], "method": [148], "substantially": [149], "surpasses": [150], "prior": [151], "baselines,": [152], "delivering": [153], "gains": [154], "up": [156], "6.0%": [158], "on": [159, 165], "mathematical": [160], "tasks": [162], "10.2%": [164], "commonsense": [166], "benchmarks.": [168]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.9023000001907349, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.9023000001907349, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.03970000147819519, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.026799999177455902, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.1109/tnnls.2025.3644134", "is_oa": false, "landing_page_url": "https://doi.org/10.1109/tnnls.2025.3644134", "pdf_url": null, "source": {"id": "https://openalex.org/S4210175523", "display_name": "IEEE Transactions on Neural Networks and Learning Systems", "issn_l": "2162-237X", "issn": ["2162-237X", "2162-2388"], "is_oa": false, "is_in_doaj": false, "is_core": true, "listed_in": ["cwts-core", "doyens", "jufo-3", "medline", "norway-2"], "host_organization": "https://openalex.org/P4310319808", "host_organization_name": "Institute of Electrical and Electronics Engineers", "host_organization_lineage": ["https://openalex.org/P4310319808"], "host_organization_lineage_names": ["Institute of Electrical and Electronics Engineers"], "type": "journal"}, "license": null, "license_id": null, "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "IEEE Transactions on Neural Networks and Learning Systems", "raw_type": "journal-article"}, "cited_by_count": 1, "is_retracted": false}, {"id": "https://openalex.org/W4415194481", "doi": "https://doi.org/10.48550/arxiv.2508.05592", "title": "MathSmith: Towards Extremely Hard Mathematical Reasoning by Forging Synthetic Problems with a Reinforced Policy", "publication_year": 2025, "publication_date": "2025-08-07", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Zhan, Shaoxiong", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I99065089", "display_name": "Tsinghua University", "ror": "https://ror.org/03cve4549", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I99065089"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Zhan, Shaoxiong", "raw_affiliation_strings": ["Tsinghua University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Tsinghua University", "institution_ids": ["https://openalex.org/I99065089"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Lai, Yanlin", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I99065089", "display_name": "Tsinghua University", "ror": "https://ror.org/03cve4549", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I99065089"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Lai, Yanlin", "raw_affiliation_strings": ["Tsinghua University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Tsinghua University", "institution_ids": ["https://openalex.org/I99065089"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5112296533", "display_name": "Ziyu Lu", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I99065089", "display_name": "Tsinghua University", "ror": "https://ror.org/03cve4549", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I99065089"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Lu, Ziyu", "raw_affiliation_strings": ["Tsinghua University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Tsinghua University", "institution_ids": ["https://openalex.org/I99065089"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5010087030", "display_name": "Dahua Lin", "orcid": "https://orcid.org/0000-0002-8865-7896", "observed_orcids": ["https://orcid.org/0000-0002-8865-7896"]}, "institutions": [{"id": "https://openalex.org/I177725633", "display_name": "Chinese University of Hong Kong", "ror": "https://ror.org/00t33hh48", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I177725633"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Lin, Dahua", "raw_affiliation_strings": ["The Chinese University of Hong Kong"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "The Chinese University of Hong Kong", "institution_ids": ["https://openalex.org/I177725633"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5001241589", "display_name": "Ziqing Yang", "orcid": "https://orcid.org/0000-0002-0492-0862", "observed_orcids": ["https://orcid.org/0000-0002-0492-0862", "https://orcid.org/0000-0001-8823-575X"]}, "institutions": [{"id": "https://openalex.org/I4401727009", "display_name": "Sensetime (China)", "ror": "https://ror.org/049wsmj07", "country_code": "CN", "type": "company", "lineage": ["https://openalex.org/I4401727009"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Yang, Ziqing", "raw_affiliation_strings": ["SenseTime Research"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "SenseTime Research", "institution_ids": ["https://openalex.org/I4401727009"]}]}, {"author_position": "last", "author": {"id": null, "display_name": "Tan, Fei", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I66867065", "display_name": "East China Normal University", "ror": "https://ror.org/02n96ep67", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I66867065"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Tan, Fei", "raw_affiliation_strings": ["East China Normal University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "East China Normal University", "institution_ids": ["https://openalex.org/I66867065"]}]}], "abstract_inverted_index": {"Large": [0], "language": [1], "models": [2], "have": [3], "achieved": [4], "substantial": [5], "progress": [6], "in": [7, 188], "mathematical": [8, 46], "reasoning,": [9], "yet": [10], "their": [11], "advancement": [12], "is": [13, 114], "limited": [14], "by": [15, 63], "the": [16, 107, 121, 182], "scarcity": [17], "of": [18, 106, 123, 184], "high-quality,": [19], "high-difficulty": [20, 185], "training": [21], "data.": [22], "Existing": [23], "synthesis": [24], "methods": [25], "largely": [26], "rely": [27], "on": [28, 170], "transforming": [29], "human-written": [30], "templates,": [31], "limiting": [32], "both": [33, 155], "diversity": [34], "and": [35, 73, 101, 142, 157, 179, 195], "scalability.": [36], "We": [37, 89], "propose": [38], "MathSmith,": [39], "a": [40, 162], "novel": [41], "framework": [42], "for": [43], "synthesizing": [44], "challenging": [45], "problems": [47, 126], "to": [48, 94, 116], "enhance": [49], "LLM": [50, 190], "reasoning.": [51, 130], "Rather": [52], "than": [53], "modifying": [54], "existing": [55, 152], "problems,": [56], "MathSmith": [57, 149, 174], "constructs": [58], "new": [59], "ones": [60], "from": [61, 68], "scratch": [62], "randomly": [64], "sampling": [65], "concept-explanation": [66], "pairs": [67], "PlanetMath,": [69], "ensuring": [70], "data": [71, 187, 196], "independence": [72], "avoiding": [74], "contamination.": [75], "To": [76], "increase": [77], "difficulty,": [78], "we": [79], "design": [80], "nine": [81], "predefined": [82], "strategies": [83], "as": [84, 136], "soft": [85], "constraints": [86], "during": [87], "rationales.": [88], "further": [90], "adopts": [91], "reinforcement": [92], "learning": [93], "jointly": [95], "optimize": [96], "structural": [97], "validity,": [98], "reasoning": [99, 108, 191], "complexity,": [100, 119], "answer": [102], "consistency.": [103], "The": [104], "length": [105], "trace": [109], "generated": [110], "under": [111, 154], "autoregressive": [112], "prompting": [113], "used": [115], "reflect": [117], "cognitive": [118], "encouraging": [120], "creation": [122], "more": [124], "demanding": [125], "aligned": [127], "with": [128], "long-chain-of-thought": [129], "Experiments": [131], "across": [132], "five": [133], "benchmarks,": [134], "categorized": [135], "easy": [137], "&": [138], "medium": [139], "(GSM8K,": [140], "MATH-500)": [141], "hard": [143], "(AIME2024,": [144], "AIME2025,": [145], "OlympiadBench),": [146], "show": [147], "that": [148], "consistently": [150], "outperforms": [151], "baselines": [153], "short": [156], "long": [158], "CoT": [159], "settings.": [160], "Additionally,": [161], "weakness-focused": [163], "variant": [164], "generation": [165], "module": [166], "enables": [167], "targeted": [168], "improvement": [169], "specific": [171], "concepts.": [172], "Overall,": [173], "exhibits": [175], "strong": [176], "scalability,": [177], "generalization,": [178], "transferability,": [180], "highlighting": [181], "promise": [183], "synthetic": [186], "advancing": [189], "capabilities.": [192], "Our": [193], "code": [194], "are": [197], "available": [198], "at": [199], "https://github.com/Jasaxion/MathSmith.": [200]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5131999850273132, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5131999850273132, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.11739999800920486, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.10589999705553055, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2508.05592", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2508.05592", "pdf_url": "https://arxiv.org/pdf/2508.05592", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415097709", "doi": "https://doi.org/10.48550/arxiv.2510.07242", "title": "Hybrid Reinforcement: When Reward Is Sparse, It's Better to Be Dense", "publication_year": 2025, "publication_date": "2025-10-08", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5057628501", "display_name": "Leitian Tao", "orcid": "https://orcid.org/0000-0002-8726-6682", "observed_orcids": ["https://orcid.org/0000-0002-8726-6682"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Tao, Leitian", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5103110668", "display_name": "Ilia Kulikov", "orcid": "https://orcid.org/0000-0001-5989-9733", "observed_orcids": ["https://orcid.org/0000-0001-5989-9733"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kulikov, Ilia", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5057314288", "display_name": "Swarnadeep Saha", "orcid": "https://orcid.org/0000-0002-6972-3448", "observed_orcids": ["https://orcid.org/0000-0002-6972-3448"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Saha, Swarnadeep", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5049397748", "display_name": "Tianlu Wang", "orcid": "https://orcid.org/0000-0002-9732-3441", "observed_orcids": ["https://orcid.org/0000-0002-9732-3441"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Tianlu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Xu, Jing", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Jing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5000775009", "display_name": "Yixuan Li", "orcid": "https://orcid.org/0009-0007-6862-9605", "observed_orcids": ["https://orcid.org/0009-0007-6862-9605"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Sharon", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5076635608", "display_name": "Jason Weston", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Weston, Jason E", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100623508", "display_name": "Ping Yu", "orcid": "https://orcid.org/0000-0002-8913-873X", "observed_orcids": ["https://orcid.org/0000-0002-8913-873X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yu, Ping", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Post-training": [0], "for": [1], "reasoning": [2, 114], "of": [3, 142, 148], "large": [4], "language": [5], "models": [6, 45, 150], "(LLMs)": [7], "increasingly": [8], "relies": [9], "on": [10, 126], "verifiable": [11, 128], "rewards:": [12], "deterministic": [13], "checkers": [14], "that": [15, 34, 71, 135], "provide": [16], "0-1": [17], "correctness": [18, 94], "signals.": [19], "While": [20], "reliable,": [21], "such": [22], "binary": [23], "feedback": [24], "is": [25], "brittle--many": [26], "tasks": [27], "admit": [28], "partially": [29], "correct": [30], "or": [31], "alternative": [32], "answers": [33], "verifiers": [35, 143], "under-credit,": [36], "and": [37, 99, 120, 129], "the": [38, 140, 146], "resulting": [39], "all-or-nothing": [40], "supervision": [41], "limits": [42], "learning.": [43], "Reward": [44, 65], "offer": [46], "richer,": [47], "continuous": [48], "feedback,": [49], "which": [50], "can": [51], "serve": [52], "as": [53], "a": [54, 67, 79], "complementary": [55], "supervisory": [56], "signal": [57], "to": [58, 86, 102, 151], "verifiers.": [59], "We": [60], "introduce": [61], "HERO": [62, 82, 116], "(Hybrid": [63], "Ensemble": [64], "Optimization),": [66], "reinforcement": [68], "learning": [69], "framework": [70], "integrates": [72], "verifier": [73], "signals": [74, 108], "with": [75, 123], "reward-model": [76, 88], "scores": [77, 89], "in": [78], "structured": [80], "way.": [81], "employs": [83], "stratified": [84], "normalization": [85], "bound": [87], "within": [90], "verifier-defined": [91], "groups,": [92], "preserving": [93], "while": [95, 144], "refining": [96], "quality": [97], "distinctions,": [98], "variance-aware": [100], "weighting": [101], "emphasize": [103], "challenging": [104], "prompts": [105], "where": [106], "dense": [107], "matter": [109], "most.": [110], "Across": [111], "diverse": [112], "mathematical": [113], "benchmarks,": [115], "consistently": [117], "outperforms": [118], "RM-only": [119], "verifier-only": [121], "baselines,": [122], "strong": [124], "gains": [125], "both": [127], "hard-to-verify": [130], "tasks.": [131], "Our": [132], "results": [133], "show": [134], "hybrid": [136], "reward": [137, 149], "design": [138], "retains": [139], "stability": [141], "leveraging": [145], "nuance": [147], "advance": [152], "reasoning.": [153]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.4553000032901764, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.4553000032901764, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.11180000007152557, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.09480000287294388, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.07242", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.07242", "pdf_url": "https://arxiv.org/pdf/2510.07242", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4406880755", "doi": "https://doi.org/10.48550/arxiv.2501.15581", "title": "Error Classification of Large Language Models on Math Word Problems: A Dynamically Adaptive Framework", "publication_year": 2025, "publication_date": "2025-01-26", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5100572120", "display_name": "Yuhong Sun", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I66867065", "display_name": "East China Normal University", "ror": "https://ror.org/02n96ep67", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I66867065"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Sun, Yuhong", "raw_affiliation_strings": ["\u2661 Software Engineering Institute , East China Normal University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "\u2661 Software Engineering Institute , East China Normal University", "institution_ids": ["https://openalex.org/I66867065"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5083955516", "display_name": "Zhangyue Yin", "orcid": "https://orcid.org/0009-0008-0620-8271", "observed_orcids": ["https://orcid.org/0009-0008-0620-8271"]}, "institutions": [{"id": "https://openalex.org/I24943067", "display_name": "Fudan University", "ror": "https://ror.org/013q1eq08", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I24943067"]}, {"id": "https://openalex.org/I4210139618", "display_name": "Shanghai Key Laboratory of Trustworthy Computing", "ror": "https://ror.org/030qbr085", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4210139618"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Yin, Zhangyue", "raw_affiliation_strings": ["School of Computer Science and Artificial Intelligence , Fudan University \u2660 Shanghai Key Laboratory of Trustworthy Computing , Shanghai , China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "School of Computer Science and Artificial Intelligence , Fudan University \u2660 Shanghai Key Laboratory of Trustworthy Computing , Shanghai , China", "institution_ids": ["https://openalex.org/I24943067", "https://openalex.org/I4210139618"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5088834359", "display_name": "Xuanjing Huang", "orcid": "https://orcid.org/0000-0001-9197-9426", "observed_orcids": ["https://orcid.org/0000-0001-9197-9426"]}, "institutions": [{"id": "https://openalex.org/I66867065", "display_name": "East China Normal University", "ror": "https://ror.org/02n96ep67", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I66867065"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Huang, Xuanjing", "raw_affiliation_strings": ["\u2661 Software Engineering Institute , East China Normal University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "\u2661 Software Engineering Institute , East China Normal University", "institution_ids": ["https://openalex.org/I66867065"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5044665993", "display_name": "Xipeng Qiu", "orcid": "https://orcid.org/0000-0001-7163-5247", "observed_orcids": ["https://orcid.org/0000-0001-7163-5247"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Qiu, Xipeng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5063868428", "display_name": "Zhao Hui", "orcid": "https://orcid.org/0000-0002-7512-3166", "observed_orcids": ["https://orcid.org/0000-0002-7512-3166"]}, "institutions": [{"id": "https://openalex.org/I66867065", "display_name": "East China Normal University", "ror": "https://ror.org/02n96ep67", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I66867065"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Zhao, Hui", "raw_affiliation_strings": ["\u2661 Software Engineering Institute , East China Normal University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "\u2661 Software Engineering Institute , East China Normal University", "institution_ids": ["https://openalex.org/I66867065"]}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "have": [4], "demonstrated": [5], "remarkable": [6], "capabilities": [7, 160], "across": [8, 85], "various": [9], "domains.": [10], "Math": [11], "Word": [12], "Problems": [13], "(MWPs)": [14], "serve": [15], "as": [16, 158, 178], "a": [17, 102, 130], "crucial": [18], "benchmark": [19], "for": [20, 133], "evaluating": [21], "LLMs'": [22], "reasoning": [23, 115, 187], "abilities.": [24], "While": [25], "most": [26], "research": [27], "primarily": [28], "focuses": [29], "on": [30, 49, 95], "improving": [31], "accuracy,": [32], "it": [33], "often": [34], "neglects": [35], "understanding": [36], "and": [37, 51, 114, 121], "addressing": [38], "the": [39, 60], "underlying": [40], "patterns": [41, 65, 113, 177], "of": [42, 63, 82, 125], "errors.": [43], "Current": [44], "error": [45, 64, 72, 76, 107, 112, 126, 136, 149, 166, 176], "classification": [46, 137], "methods": [47], "rely": [48], "static": [50], "predefined": [52], "categories,": [53], "which": [54, 151], "limit": [55], "their": [56], "ability": [57], "to": [58, 155, 182], "capture": [59], "full": [61], "spectrum": [62], "in": [66, 138, 185], "mathematical": [67, 139, 186], "reasoning.": [68, 140], "To": [69, 117], "enable": [70, 122], "systematic": [71], "analysis,": [73], "we": [74, 99, 128, 168], "collect": [75], "samples": [77, 108], "from": [78, 153], "15": [79], "different": [80], "LLMs": [81], "varying": [83], "sizes": [84], "four": [86], "distinct": [87], "MWP": [88], "datasets": [89], "using": [90], "multiple": [91], "sampling": [92], "strategies.": [93], "Based": [94], "this": [96], "extensive": [97], "collection,": [98], "introduce": [100], "MWPES-300K,": [101], "comprehensive": [103], "dataset": [104, 145], "containing": [105], "304,865": [106], "that": [109, 144, 173], "cover": [110], "diverse": [111], "paths.": [116], "reduce": [118], "human": [119], "bias": [120], "fine-grained": [123], "analysis": [124], "patterns,": [127, 150, 167], "propose": [129, 169], "novel": [131], "framework": [132], "automated": [134], "dynamic": [135], "Experimental": [141], "results": [142], "demonstrate": [143], "characteristics": [146], "significantly": [147], "shape": [148], "evolve": [152], "basic": [154], "complex": [156], "manifestations": [157], "model": [159], "increase.": [161], "With": [162], "deeper": [163], "insights": [164], "into": [165], "Error-Aware": [170], "Prompting": [171], "(EAP)": [172], "incorporates": [174], "common": [175], "explicit": [179], "guidance,": [180], "leading": [181], "significant": [183], "improvements": [184], "performance.": [188]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7226999998092651, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7226999998092651, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.07590000331401825, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.06289999932050705, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2501.15581", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2501.15581", "pdf_url": "https://arxiv.org/pdf/2501.15581", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by-sa", "license_id": "https://openalex.org/licenses/cc-by-sa", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 1, "is_retracted": false}, {"id": "https://openalex.org/W7117475173", "doi": "https://doi.org/10.36227/techrxiv.176704492.22222926/v1", "title": "Tool-Augmented Hybrid Ensemble Reasoning with Distillation for Bilingual Mathematical Problem Solving", "publication_year": 2025, "publication_date": "2025-12-29", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5100563441", "display_name": "Peiqing Lu", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I111088046", "display_name": "Boston University", "ror": "https://ror.org/05qwgg493", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I111088046"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Peiqing Lu", "raw_affiliation_strings": ["Boston University Boston"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Boston University Boston", "institution_ids": ["https://openalex.org/I111088046"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5146599230", "display_name": "Yuan Zhang", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I111088046", "display_name": "Boston University", "ror": "https://ror.org/05qwgg493", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I111088046"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Yuan Zhang", "raw_affiliation_strings": ["Boston University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Boston University", "institution_ids": ["https://openalex.org/I111088046"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101724788", "display_name": "Haoyun Zhang", "orcid": "https://orcid.org/0000-0002-2666-6850", "observed_orcids": ["https://orcid.org/0000-0002-2666-6850"]}, "institutions": [{"id": "https://openalex.org/I79576946", "display_name": "University of Pennsylvania", "ror": "https://ror.org/00b30xv10", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I79576946"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Haoyun Zhang", "raw_affiliation_strings": ["University of Pennsylvania"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University of Pennsylvania", "institution_ids": ["https://openalex.org/I79576946"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5061351647", "display_name": "Jiasen Zheng", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I111979921", "display_name": "Northwestern University", "ror": "https://ror.org/000e0be47", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I111979921"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Jiasen Zheng", "raw_affiliation_strings": ["Northwestern University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Northwestern University", "institution_ids": ["https://openalex.org/I111979921"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121517005", "display_name": "Wenjun Wu", "orcid": "https://orcid.org/0000-0002-6413-5700", "observed_orcids": ["https://orcid.org/0000-0002-6413-5700"]}, "institutions": [{"id": "https://openalex.org/I157725225", "display_name": "University of Illinois Urbana-Champaign", "ror": "https://ror.org/047426m28", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I157725225"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Wenjun Wu", "raw_affiliation_strings": ["University of Illinois Urbana-Champaign Urbana"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University of Illinois Urbana-Champaign Urbana", "institution_ids": ["https://openalex.org/I157725225"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5121204086", "display_name": "Kejian Tong", "orcid": "https://orcid.org/0009-0002-5127-2711", "observed_orcids": ["https://orcid.org/0009-0002-5127-2711"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kejian Tong", "raw_affiliation_strings": [], "raw_orcid": "https://orcid.org/0009-0002-5127-2711", "affiliations": []}], "abstract_inverted_index": {"Bilingual": [0], "mathematical": [1, 119], "problem": [2], "solving": [3], "needs": [4], "a": [5, 39, 114], "clear": [6], "link": [7], "between": [8], "language": [9, 15, 19], "reasoning": [10, 43, 64, 108, 120], "and": [11, 37, 44, 49, 58, 71, 85, 101, 109, 125], "symbolic": [12, 97], "calculation.": [13, 111], "Large": [14], "models": [16], "often": [17], "handle": [18], "well": [20], "but": [21], "are": [22], "weak": [23], "in": [24], "accurate": [25], "computation.": [26], "This": [27], "paper": [28], "presents": [29], "HERALD": [30, 51, 112], "(Hybrid": [31], "Ensemble": [32], "Reasoning": [33], "with": [34, 121], "Adaptive": [35], "Learning": [36], "Distillation),": [38], "framework": [40], "that": [41, 95], "joins": [42], "calculation": [45], "using": [46], "NuminaMath-7B-TIR,": [47], "GPT-4o,": [48], "Mistral-7B.": [50], "uses": [52], "adaptive": [53, 99], "routing,": [54], "tool-based": [55], "reinforcement": [56], "learning,": [57], "knowledge": [59], "distillation": [60, 86], "to": [61, 82], "connect": [62], "different": [63], "paths.": [65], "Confidence": [66], "calibration": [67], "keeps": [68, 74], "weighting": [69], "stable,": [70], "dual-path": [72], "checking": [73], "results": [75], "correct.": [76], "Reinforcement": [77], "learning": [78], "controls": [79], "tool": [80], "use": [81], "cut": [83], "redundancy,": [84], "lowers": [87], "delay": [88], "without": [89], "hurting": [90], "accuracy.": [91], "The": [92], "system": [93], "shows": [94], "combining": [96], "checking,": [98], "ensembles,": [100], "bilingual": [102], "fine-tuning": [103], "helps": [104], "achieve": [105], "both": [106], "fluent": [107], "precise": [110], "offers": [113], "practical": [115], "solution": [116], "for": [117], "multilingual": [118], "better": [122], "accuracy,": [123], "stability,": [124], "clarity.": [126]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.2919999957084656, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.2919999957084656, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.2542000114917755, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.11749999970197678, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.36227/techrxiv.176704492.22222926/v1", "is_oa": true, "landing_page_url": "https://doi.org/10.36227/techrxiv.176704492.22222926/v1", "pdf_url": null, "source": null, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "acceptedVersion", "is_accepted": true, "is_published": false, "raw_source_name": null, "raw_type": "posted-content"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414900373", "doi": "https://doi.org/10.48550/arxiv.2503.12854", "title": "Enhancing LLM Reasoning with Iterative DPO: A Comprehensive Empirical Investigation", "publication_year": 2025, "publication_date": "2025-03-17", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5008658985", "display_name": "Songjun Tu", "orcid": "https://orcid.org/0009-0009-6609-762X", "observed_orcids": ["https://orcid.org/0009-0009-6609-762X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Tu, Songjun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Lin, Jiahao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Lin, Jiahao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5009748337", "display_name": "Xiangyu Tian", "orcid": "https://orcid.org/0000-0002-6606-5005", "observed_orcids": ["https://orcid.org/0000-0002-6606-5005"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Tian, Xiangyu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Zhang, Qichao", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Qichao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101428437", "display_name": "Linjing Li", "orcid": "https://orcid.org/0000-0001-8647-495X", "observed_orcids": ["https://orcid.org/0000-0001-8647-495X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Linjing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5072248915", "display_name": "Yuqian Fu", "orcid": "https://orcid.org/0000-0002-0412-5500", "observed_orcids": ["https://orcid.org/0000-0002-0412-5500"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Fu, Yuqian", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5025274226", "display_name": "Nan Xu", "orcid": "https://orcid.org/0000-0001-9737-396X", "observed_orcids": ["https://orcid.org/0000-0001-9737-396X", "https://orcid.org/0000-0001-5996-9201"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xu, Nan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5115597121", "display_name": "Wei He", "orcid": "https://orcid.org/0000-0002-3100-9724", "observed_orcids": ["https://orcid.org/0000-0002-3100-9724"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "He, Wei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5055543314", "display_name": "Xiangyuan Lan", "orcid": "https://orcid.org/0000-0001-8564-0346", "observed_orcids": ["https://orcid.org/0000-0001-8564-0346"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Lan, Xiangyuan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5015286125", "display_name": "Dongmei Jiang", "orcid": "https://orcid.org/0000-0002-2142-2013", "observed_orcids": ["https://orcid.org/0000-0002-2142-2013"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Jiang, Dongmei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100624298", "display_name": "Dongbin Zhao", "orcid": "https://orcid.org/0000-0001-8218-9633", "observed_orcids": ["https://orcid.org/0000-0001-8218-9633"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhao, Dongbin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Recent": [0], "advancements": [1], "in": [2, 36, 54, 148], "post-training": [3], "methodologies": [4], "for": [5, 19, 57, 80, 91, 144], "large": [6], "language": [7], "models": [8], "(LLMs)": [9], "have": [10, 31], "highlighted": [11], "reinforcement": [12], "learning": [13], "(RL)": [14], "as": [15, 40, 132], "a": [16, 66, 133, 141], "critical": [17], "component": [18], "enhancing": [20, 145], "reasoning.": [21], "However,": [22], "the": [23, 50, 93, 96], "substantial": [24], "computational": [25, 126], "costs": [26], "associated": [27], "with": [28, 71, 113, 123], "RL-based": [29], "approaches": [30], "led": [32], "to": [33, 138], "growing": [34], "interest": [35], "alternative": [37, 137], "paradigms,": [38], "such": [39], "Direct": [41], "Preference": [42], "Optimization": [43], "(DPO).": [44], "In": [45], "this": [46], "study,": [47], "we": [48, 85], "investigate": [49], "effectiveness": [51], "of": [52, 69, 110], "DPO": [53, 70, 131], "facilitating": [55], "self-improvement": [56], "LLMs": [58], "through": [59, 104], "iterative": [60, 88], "preference-based": [61], "learning.": [62], "We": [63], "demonstrate": [64], "that": [65], "single": [67], "round": [68], "coarse": [72], "filtering": [73], "significantly": [74, 124], "enhances": [75], "mathematical": [76], "reasoning": [77, 147], "performance,": [78], "particularly": [79], "strong": [81], "base": [82], "model.": [83], "Furthermore,": [84], "design": [86], "an": [87], "enhancement": [89], "framework": [90], "both": [92], "generator": [94], "and": [95, 135], "reward": [97], "model": [98, 118], "(RM),": [99], "enabling": [100], "their": [101], "mutual": [102], "improvement": [103], "online": [105], "interaction": [106], "across": [107], "multiple": [108], "rounds": [109], "DPO.": [111], "Finally,": [112], "simple": [114], "verifiable": [115], "rewards,": [116], "our": [117], "DPO-VP": [119], "achieves": [120], "RL-level": [121], "performance": [122], "lower": [125], "overhead.": [127], "These": [128], "findings": [129], "highlight": [130], "scalable": [134], "cost-effective": [136], "RL,": [139], "offering": [140], "practical": [142], "solution": [143], "LLM": [146], "resource-constrained": [149], "situations.": [150]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.38960000872612, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.38960000872612, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.14949999749660492, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.133200004696846, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2503.12854", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2503.12854", "pdf_url": "https://arxiv.org/pdf/2503.12854", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7108666296", "doi": "https://doi.org/10.5281/zenodo.17814910", "title": "The Shadow Simplex: A Critical Framework for Analyzing Failure Modes in Self-Evolving Multi-Agent Reinforcement Learning Systems", "publication_year": 2025, "publication_date": "2025-12-04", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Poschel, Timothy", "orcid": "https://orcid.org/0009-0008-0265-9761", "observed_orcids": ["https://orcid.org/0009-0008-0265-9761"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Poschel, Timothy", "raw_affiliation_strings": [], "raw_orcid": "https://orcid.org/0009-0008-0265-9761", "affiliations": []}], "abstract_inverted_index": {"Self-evolving": [0], "agent": [1], "systems": [2, 26, 169], "that": [3, 112], "operate": [4], "without": [5], "external": [6], "data": [7], "represent": [8], "a": [9, 42, 61, 90, 163, 171], "paradigm": [10], "shift": [11], "in": [12, 32, 51, 107, 117, 123], "machine": [13], "learning,": [14], "promising": [15], "to": [16, 102, 134, 156], "transcend": [17], "the": [18, 38, 140], "limitations": [19], "of": [20], "human-annotated": [21], "datasets.": [22], "However,": [23], "these": [24, 104], "closed-loop": [25], "introduce": [27], "novel": [28], "failure": [29, 105], "modes": [30, 106], "absent": [31], "traditional": [33], "supervised": [34], "learning.": [35], "I": [36, 64, 93, 138], "present": [37], "Shadow": [39], "Simplex": [40], "Framework,": [41], "mathematical": [43, 118], "and": [44, 76, 84, 120, 132, 149, 153, 170], "conceptual": [45], "tool": [46, 70, 165], "for": [47, 166, 174], "analyzing": [48], "pathological": [49], "attractors": [50], "self-evolving": [52, 168], "systems.": [53], "Using": [54], "Agent0": [55], "(Xia": [56], "et": [57], "al.,": [58], "2025)": [59], "as": [60], "case": [62], "study,": [63], "identify": [65], "five": [66], "fundamental": [67], "pathologies\u2014co-evolutionary": [68], "collapse,": [69], "dependency,": [71], "context": [72], "corruption,": [73], "reward": [74], "hacking,": [75], "optimization": [77], "instability\u2014and": [78], "map": [79], "their": [80, 176], "10": [81, 85], "pairwise": [82], "couplings": [83], "higher-order": [86], "emergent": [87], "dysfunctions": [88], "onto": [89], "4-simplex": [91], "topology.": [92], "propose": [94, 154], "35": [95], "testable": [96], "hypotheses": [97], "with": [98], "concrete": [99], "experimental": [100], "protocols": [101], "detect": [103], "practice.": [108], "My": [109], "analysis": [110], "reveals": [111], "Agent0's": [113], "reported": [114, 147], "18%": [115], "improvement": [116], "reasoning": [119, 125], "24%": [121], "gain": [122], "general": [124], "may": [126], "mask": [127], "underlying": [128], "brittleness,": [129], "metric": [130], "manipulation,": [131], "convergence": [133], "shallow": [135], "local": [136], "optima.": [137], "validate": [139], "framework": [141, 173], "by": [142], "demonstrating": [143], "internal": [144], "inconsistencies": [145], "between": [146], "metrics": [148], "expected": [150], "theoretical": [151], "behaviors,": [152], "modifications": [155], "improve": [157], "robustness.": [158], "This": [159], "work": [160], "contributes": [161], "both": [162], "diagnostic": [164], "evaluating": [167], "cautionary": [172], "interpreting": [175], "benchmarks.": [177]}, "primary_topic": {"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.7567999958992004, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10462", "display_name": "Reinforcement Learning in Robotics", "score": 0.7567999958992004, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11975", "display_name": "Evolutionary Algorithms and Applications", "score": 0.13349999487400055, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T14381", "display_name": "Psychiatry, Mental Health, Neuroscience", "score": 0.04100000113248825, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5281/zenodo.17814910", "is_oa": true, "landing_page_url": "https://doi.org/10.5281/zenodo.17814910", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400562", "display_name": "Zenodo (CERN European Organization for Nuclear Research)", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I67311998", "host_organization_name": "European Organization for Nuclear Research", "host_organization_lineage": ["https://openalex.org/I67311998"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "Preprint"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417113157", "doi": "https://doi.org/10.48550/arxiv.2510.08595", "title": "Systematic Diagnosis of Brittle Reasoning in Large Language Models", "publication_year": 2025, "publication_date": "2025-10-05", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5120720828", "display_name": "V. S. Raghu Parupudi", "orcid": "https://orcid.org/0009-0001-5888-4342", "observed_orcids": ["https://orcid.org/0009-0001-5888-4342"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Parupudi, V. S. Raghu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"A": [0], "central": [1], "question": [2], "in": [3], "artificial": [4], "intelligence": [5], "is": [6], "the": [7, 48, 90, 115], "extent": [8], "to": [9, 33, 60, 73, 129], "which": [10], "machine": [11], "learning": [12], "models": [13], "comprehend": [14], "mathematics.": [15], "To": [16], "address": [17], "this,": [18], "we": [19], "propose": [20], "a": [21, 54, 81, 85, 125, 135], "novel": [22], "framework": [23], "for": [24, 138], "measuring": [25], "mathematical": [26, 131], "reasoning": [27, 44, 71, 107, 120], "that": [28], "moves": [29], "beyond": [30], "standard": [31], "benchmarks": [32], "diagnose": [34], "specific": [35], "failure": [36], "points.": [37], "Our": [38], "method": [39, 128], "first": [40], "generates": [41], "structured,": [42], "step-by-step": [43], "from": [45], "gpt-3.5-turbo": [46], "on": [47, 95, 103], "GSM8K": [49], "dataset.": [50], "We": [51], "then": [52], "use": [53], "more": [55, 126, 143], "capable": [56], "analyst": [57], "model,": [58], "gpt-4o-mini,": [59], "categorize": [61], "errors": [62], "and,": [63], "crucially,": [64], "perform": [65], "an": [66], "unsupervised": [67], "clustering": [68], "of": [69, 117], "every": [70], "sentence": [72], "identify": [74], "emergent": [75], "\"reasoning": [76], "modes.\"": [77], "This": [78], "analysis": [79], "reveals": [80], "cognitive": [82], "profile": [83], "with": [84, 108], "stark,": [86], "nonhuman-like": [87], "brittleness:": [88], "while": [89], "model": [91], "achieves": [92], "near-perfect": [93], "accuracy": [94], "procedural": [96], "modes": [97, 104], "like": [98], "sequential": [99], "calculation,": [100], "its": [101], "performance": [102], "requiring": [105], "combinatorial": [106], "restrictions": [109], "plummets.": [110], "By": [111], "identifying": [112], "and": [113, 133, 142], "quantifying": [114], "reliability": [116], "these": [118], "distinct": [119], "skills,": [121], "our": [122], "work": [123], "provides": [124], "granular": [127], "evaluate": [130], "comprehension": [132], "offers": [134], "precise": [136], "roadmap": [137], "developing": [139], "new": [140], "capabilities": [141], "reliable": [144], "future": [145], "applications.": [146]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.39500001072883606, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.39500001072883606, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.094200000166893, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.0697999969124794, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.08595", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.08595", "pdf_url": "https://arxiv.org/pdf/2510.08595", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4411824682", "doi": "https://doi.org/10.1007/s11071-025-11522-4", "title": "Using the bilinear neural network method and the neural network-based symbolic calculation approach to find exact solutions of the (3+1)-dimensional Kairat-X extended equation", "publication_year": 2025, "publication_date": "2025-06-30", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Ni Hai", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I55654194", "display_name": "Inner Mongolia University of Technology", "ror": "https://ror.org/05564e019", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I55654194"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Ni Hai", "raw_affiliation_strings": ["College of Sciences, Inner Mongolia University of Technology, Hohhot, 010051, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "College of Sciences, Inner Mongolia University of Technology, Hohhot, 010051, China", "institution_ids": ["https://openalex.org/I55654194"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5049378399", "display_name": "Sudao Bilige", "orcid": "https://orcid.org/0000-0003-3546-6751", "observed_orcids": ["https://orcid.org/0000-0003-3546-6751"]}, "institutions": [{"id": "https://openalex.org/I55654194", "display_name": "Inner Mongolia University of Technology", "ror": "https://ror.org/05564e019", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I55654194"]}], "countries": ["CN"], "is_corresponding": true, "raw_author_name": "Sudao Bilige", "raw_affiliation_strings": ["College of Sciences, Inner Mongolia University of Technology, Hohhot, 010051, China", "Inner Mongolia Key Laboratory of Statistical Analysis Theory for Life Data and Neural Network Modeling, Hohhot, 010051, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "College of Sciences, Inner Mongolia University of Technology, Hohhot, 010051, China", "institution_ids": ["https://openalex.org/I55654194"]}, {"raw_affiliation_string": "Inner Mongolia Key Laboratory of Statistical Analysis Theory for Life Data and Neural Network Modeling, Hohhot, 010051, China", "institution_ids": []}]}], "abstract_inverted_index": null, "primary_topic": {"id": "https://openalex.org/T10248", "display_name": "Nonlinear Waves and Solitons", "score": 0.9829000234603882, "subfield": {"id": "https://openalex.org/subfields/3109", "display_name": "Statistical and Nonlinear Physics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10248", "display_name": "Nonlinear Waves and Solitons", "score": 0.9829000234603882, "subfield": {"id": "https://openalex.org/subfields/3109", "display_name": "Statistical and Nonlinear Physics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11206", "display_name": "Model Reduction and Neural Networks", "score": 0.01080000028014183, "subfield": {"id": "https://openalex.org/subfields/3109", "display_name": "Statistical and Nonlinear Physics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10320", "display_name": "Neural Networks and Applications", "score": 0.002199999988079071, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.1007/s11071-025-11522-4", "is_oa": false, "landing_page_url": "https://doi.org/10.1007/s11071-025-11522-4", "pdf_url": null, "source": {"id": "https://openalex.org/S138681734", "display_name": "Nonlinear Dynamics", "issn_l": "0924-090X", "issn": ["0924-090X", "1573-269X"], "is_oa": false, "is_in_doaj": false, "is_core": true, "listed_in": ["cwts-core", "jufo-2", "norway-1"], "host_organization": "https://openalex.org/P4310319900", "host_organization_name": "Springer Science+Business Media", "host_organization_lineage": ["https://openalex.org/P4310319900", "https://openalex.org/P4310319965"], "host_organization_lineage_names": ["Springer Science+Business Media", "Springer Nature"], "type": "journal"}, "license": null, "license_id": null, "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Nonlinear Dynamics", "raw_type": "journal-article"}, "cited_by_count": 10, "is_retracted": false}, {"id": "https://openalex.org/W4416811821", "doi": "https://doi.org/10.18061/ojsm.6674", "title": "From Prompt to Practice: Using Artificial Intelligence to Support Productive Struggle in Elementary Mathematics Methods Courses", "publication_year": 2025, "publication_date": "2025-11-29", "type": "article", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Alesia Mickle Moldavan", "orcid": "https://orcid.org/0000-0001-7946-9028", "observed_orcids": ["https://orcid.org/0000-0001-7946-9028"]}, "institutions": [{"id": "https://openalex.org/I39815113", "display_name": "Georgia Southern University", "ror": "https://ror.org/04agmb972", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I39815113"]}], "countries": ["US"], "is_corresponding": true, "raw_author_name": "Alesia Mickle Moldavan", "raw_affiliation_strings": ["Georgia Southern University"], "raw_orcid": "https://orcid.org/0000-0001-7946-9028", "affiliations": [{"raw_affiliation_string": "Georgia Southern University", "institution_ids": ["https://openalex.org/I39815113"]}]}], "abstract_inverted_index": {"This": [0], "study": [1], "investigates": [2], "the": [3, 17, 102], "use": [4], "of": [5, 94, 114, 129], "AI": [6, 73, 103, 162], "to": [7, 24, 75, 169], "support": [8], "elementary": [9, 40], "teachers": [10, 56, 67], "in": [11, 142], "developing": [12], "mathematical": [13, 33, 83, 115, 145], "story": [14], "problems,": [15], "alongside": [16], "content": [18], "knowledge": [19], "and": [20, 32, 60, 89, 98, 117, 123, 133, 137, 155, 166], "pedagogical": [21], "strategies": [22], "needed": [23], "facilitate": [25], "problem-solving": [26, 131], "that": [27], "encourages": [28], "critical": [29], "thinking,": [30], "perseverance,": [31], "discussion.": [34], "Specifically,": [35], "MagicSchool.ai": [36], "was": [37], "integrated": [38], "into": [39, 163], "mathematics": [41, 173], "methods": [42], "courses": [43], "at": [44], "a": [45, 152], "southeastern": [46], "United": [47], "States": [48], "university:": [49], "two": [50, 61], "asynchronous": [51], "graduate": [52], "sections": [53, 64], "for": [54, 65, 135, 160], "in-service": [55], "(n": [57, 68], "=": [58, 69], "31)": [59], "in-person": [62], "undergraduate": [63], "preservice": [66], "46).": [70], "Participants": [71], "used": [72], "tools": [74, 104], "examine": [76], "how": [77, 101], "their": [78, 106, 112, 127, 140], "questioning": [79], "techniques": [80], "could": [81], "enrich": [82], "reasoning": [84], "while": [85], "maintaining": [86], "cognitive": [87], "demand": [88], "promoting": [90], "productive": [91, 144], "struggle.": [92], "Analysis": [93], "participants\u2019": [95], "course": [96], "artifacts": [97], "reflections": [99], "revealed": [100], "strengthened": [105], "instructional": [107], "practice": [108], "by:": [109], "(a)": [110], "deepening": [111], "understanding": [113], "processes": [116], "scaffolding,": [118], "(b)": [119], "improving": [120], "question": [121], "types": [122], "sequencing,": [124], "(c)": [125], "broadening": [126], "exploration": [128], "diverse": [130], "approaches": [132], "opportunities": [134], "personalization,": [136], "(d)": [138], "enhancing": [139], "self-reflection": [141], "facilitating": [143], "discussions.": [146], "Findings": [147], "highlight": [148], "AI\u2019s": [149], "potential": [150], "as": [151], "prompt": [153], "generator": [154], "question-scripting": [156], "partner,": [157], "with": [158], "recommendations": [159], "integrating": [161], "teacher": [164], "preparation": [165], "K\u201312": [167], "classrooms": [168], "promote": [170], "reflective,": [171], "effective": [172], "instruction.": [174]}, "primary_topic": {"id": "https://openalex.org/T10130", "display_name": "Mathematics Education and Teaching Techniques", "score": 0.8788999915122986, "subfield": {"id": "https://openalex.org/subfields/3304", "display_name": "Education"}, "field": {"id": "https://openalex.org/fields/33", "display_name": "Social Sciences"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, "topics": [{"id": "https://openalex.org/T10130", "display_name": "Mathematics Education and Teaching Techniques", "score": 0.8788999915122986, "subfield": {"id": "https://openalex.org/subfields/3304", "display_name": "Education"}, "field": {"id": "https://openalex.org/fields/33", "display_name": "Social Sciences"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T14414", "display_name": "Artificial Intelligence in Education", "score": 0.11240000277757645, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11902", "display_name": "Intelligent Tutoring Systems and Adaptive Learning", "score": 0.0024999999441206455, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.18061/ojsm.6674", "is_oa": true, "landing_page_url": "https://doi.org/10.18061/ojsm.6674", "pdf_url": null, "source": {"id": "https://openalex.org/S5407051569", "display_name": "Ohio Journal of School Mathematics", "issn_l": "2472-5986", "issn": ["2472-5986", "2472-601X"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": null, "host_organization_name": null, "host_organization_lineage": [], "host_organization_lineage_names": [], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Ohio Journal of School Mathematics", "raw_type": "journal-article"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415032442", "doi": "https://doi.org/10.48550/arxiv.2505.02735", "title": "FormalMATH: Benchmarking Formal Mathematical Reasoning of Large Language Models", "publication_year": 2025, "publication_date": "2025-05-05", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5101353826", "display_name": "Zhouliang Yu", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I177725633", "display_name": "Chinese University of Hong Kong", "ror": "https://ror.org/00t33hh48", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I177725633"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Yu, Zhouliang", "raw_affiliation_strings": ["The Chinese University of Hong Kong"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "The Chinese University of Hong Kong", "institution_ids": ["https://openalex.org/I177725633"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5113358583", "display_name": "Ruotian Peng", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I3133055985", "display_name": "Westlake University", "ror": "https://ror.org/05hfa4n20", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I3133055985"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Peng, Ruotian", "raw_affiliation_strings": ["Westlake University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Westlake University", "institution_ids": ["https://openalex.org/I3133055985"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5113367762", "display_name": "Keyi Ding", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ding, Keyi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5045807386", "display_name": "Yizhe Li", "orcid": "https://orcid.org/0000-0002-4379-6200", "observed_orcids": ["https://orcid.org/0000-0002-4379-6200", "https://orcid.org/0009-0007-8863-978X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Yizhe", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5073605360", "display_name": "Zhongyuan Peng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Peng, Zhongyuan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5059310859", "display_name": "Minghao Liu", "orcid": "https://orcid.org/0000-0002-7906-5530", "observed_orcids": ["https://orcid.org/0000-0002-7906-5530"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Minghao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5046456999", "display_name": "Y. M. Zhang", "orcid": "https://orcid.org/0009-0005-2061-0273", "observed_orcids": ["https://orcid.org/0009-0005-2061-0273"]}, "institutions": [{"id": "https://openalex.org/I161318765", "display_name": "University of California, Los Angeles", "ror": "https://ror.org/046rm7j60", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I161318765"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Zhang, Yifan", "raw_affiliation_strings": ["University of California , Los Angeles"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University of California , Los Angeles", "institution_ids": ["https://openalex.org/I161318765"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100342744", "display_name": "Zheng Yuan", "orcid": "https://orcid.org/0009-0005-1079-6481", "observed_orcids": ["https://orcid.org/0009-0005-1079-6481"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Yuan, Zheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5111038333", "display_name": "Huajian Xin", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xin, Huajian", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5083410012", "display_name": "Wenhao Huang", "orcid": "https://orcid.org/0000-0002-0036-6278", "observed_orcids": ["https://orcid.org/0000-0002-0036-6278", "https://orcid.org/0009-0008-3041-3769"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Huang, Wenhao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5070584981", "display_name": "Yandong Wen", "orcid": "https://orcid.org/0000-0001-6330-7438", "observed_orcids": ["https://orcid.org/0000-0001-6330-7438"]}, "institutions": [{"id": "https://openalex.org/I3133055985", "display_name": "Westlake University", "ror": "https://ror.org/05hfa4n20", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I3133055985"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Wen, Yandong", "raw_affiliation_strings": ["Westlake University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Westlake University", "institution_ids": ["https://openalex.org/I3133055985"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100326077", "display_name": "Ge Zhang", "orcid": "https://orcid.org/0000-0003-0704-0657", "observed_orcids": ["https://orcid.org/0000-0003-0704-0657"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Ge", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5101606029", "display_name": "Weiyang Liu", "orcid": "https://orcid.org/0000-0002-8234-3263", "observed_orcids": ["https://orcid.org/0000-0002-8234-3263", "https://orcid.org/0009-0002-4814-0187"]}, "institutions": [{"id": "https://openalex.org/I149899117", "display_name": "Max Planck Society", "ror": "https://ror.org/01hhn8329", "country_code": "DE", "type": "nonprofit", "lineage": ["https://openalex.org/I149899117"]}, {"id": "https://openalex.org/I4210135521", "display_name": "Max Planck Institute for Intelligent Systems", "ror": "https://ror.org/04fq9j139", "country_code": "DE", "type": "facility", "lineage": ["https://openalex.org/I149899117", "https://openalex.org/I4210135521"]}], "countries": ["DE"], "is_corresponding": false, "raw_author_name": "Liu, Weiyang", "raw_affiliation_strings": ["Max Planck Institute for Intelligent Systems , T\u00fcbingen"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Max Planck Institute for Intelligent Systems , T\u00fcbingen", "institution_ids": ["https://openalex.org/I149899117", "https://openalex.org/I4210135521"]}]}], "abstract_inverted_index": {"Formal": [0], "mathematical": [1, 202], "reasoning": [2, 174, 180, 189], "remains": [3], "a": [4, 26, 65, 161, 196], "critical": [5], "challenge": [6], "for": [7, 78, 199], "artificial": [8], "intelligence,": [9], "hindered": [10], "by": [11, 101], "limitations": [12], "of": [13, 60, 104, 119], "existing": [14], "benchmarks": [15], "in": [16, 146, 150, 172, 186], "scope": [17], "and": [18, 53, 85, 152, 169], "scale.": [19], "To": [20, 56], "address": [21], "this,": [22], "we": [23, 63, 159], "present": [24], "FormalMATH,": [25], "large-scale": [27], "Lean4": [28], "benchmark": [29, 198], "comprising": [30], "5,560": [31], "formally": [32], "verified": [33], "problems": [34], "spanning": [35], "from": [36], "high-school": [37], "Olympiad": [38], "challenges": [39], "to": [40, 112], "undergraduate-level": [41], "theorems": [42], "across": [43], "diverse": [44], "domains": [45], "(e.g.,": [46, 144], "algebra,": [47], "applied": [48], "mathematics,": [49], "calculus,": [50], "number": [51], "theory,": [52], "discrete": [54], "mathematics).": [55], "mitigate": [57], "the": [58, 113, 128, 187], "inefficiency": [59], "manual": [61, 107], "formalization,": [62], "introduce": [64], "novel": [66], "human-in-the-loop": [67], "autoformalization": [68], "pipeline": [69], "that": [70, 177, 193], "integrates:": [71], "(1)": [72], "specialized": [73], "large": [74], "language": [75], "models": [76, 130], "(LLMs)": [77], "statement": [79], "autoformalization,": [80], "(2)": [81], "multi-LLM": [82], "semantic": [83], "verification,": [84], "(3)": [86], "negation-based": [87], "disproof": [88], "filtering": [89], "strategies": [90], "using": [91], "off-the-shelf": [92], "LLM-based": [93, 121], "provers.": [94], "This": [95], "approach": [96], "reduces": [97], "expert": [98], "annotation": [99], "costs": [100], "retaining": [102], "72.09%": [103], "statements": [105], "before": [106], "verification": [108], "while": [109], "ensuring": [110], "fidelity": [111], "original": [114], "natural-language": [115, 166], "problems.": [116], "Our": [117], "evaluation": [118], "state-of-the-art": [120], "theorem": [122], "provers": [123], "reveals": [124], "significant": [125], "limitations:": [126], "even": [127], "strongest": [129], "achieve": [131], "only": [132], "16.46%": [133], "success": [134, 171], "rate": [135], "under": [136], "practical": [137], "sampling": [138], "budgets,": [139], "exhibiting": [140], "pronounced": [141], "domain": [142], "bias": [143], "excelling": [145], "algebra": [147], "but": [148], "failing": [149], "calculus)": [151], "over-reliance": [153], "on": [154], "simplified": [155], "automation": [156], "tactics.": [157], "Notably,": [158], "identify": [160], "counterintuitive": [162], "inverse": [163], "relationship": [164], "between": [165], "solution": [167], "guidance": [168], "proof": [170], "chain-of-thought": [173], "scenarios,": [175], "suggesting": [176], "human-written": [178], "informal": [179], "introduces": [181], "noise": [182], "rather": [183], "than": [184], "clarity": [185], "formal": [188, 201], "settings.": [190], "We": [191], "believe": [192], "FormalMATH": [194], "provides": [195], "robust": [197], "benchmarking": [200], "reasoning.": [203]}, "primary_topic": {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.4968000054359436, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.4968000054359436, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.37209999561309814, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.021199999377131462, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2505.02735", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2505.02735", "pdf_url": "https://arxiv.org/pdf/2505.02735", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4408503389", "doi": "https://doi.org/10.21468/scipostphys.18.3.096", "title": "Generative learning of continuous data by tensor networks", "publication_year": 2025, "publication_date": "2025-03-17", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5038705915", "display_name": "Alex Meiburg", "orcid": "https://orcid.org/0000-0002-4506-9146", "observed_orcids": ["https://orcid.org/0000-0002-4506-9146"]}, "institutions": [{"id": "https://openalex.org/I183974707", "display_name": "Perimeter Institute", "ror": "https://ror.org/013m0ej23", "country_code": "CA", "type": "facility", "lineage": ["https://openalex.org/I183974707"]}, {"id": "https://openalex.org/I4210111877", "display_name": "Zapata (United States)", "ror": "https://ror.org/01bhrtv82", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I4210111877"]}], "countries": ["CA", "US"], "is_corresponding": false, "raw_author_name": "Alex Meiburg", "raw_affiliation_strings": ["Institute for Quantum Computing", "Perimeter Institute", "Zapata (United States)"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Institute for Quantum Computing", "institution_ids": []}, {"raw_affiliation_string": "Perimeter Institute", "institution_ids": ["https://openalex.org/I183974707"]}, {"raw_affiliation_string": "Zapata (United States)", "institution_ids": ["https://openalex.org/I4210111877"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100394809", "display_name": "Jing Chen", "orcid": "https://orcid.org/0000-0002-5596-4013", "observed_orcids": ["https://orcid.org/0000-0002-5596-4013", "https://orcid.org/0000-0003-0538-689X", "https://orcid.org/0000-0001-6365-902X"]}, "institutions": [{"id": "https://openalex.org/I4210111877", "display_name": "Zapata (United States)", "ror": "https://ror.org/01bhrtv82", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I4210111877"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Jing Chen", "raw_affiliation_strings": ["Zapata (United States)"], "raw_orcid": "https://orcid.org/0000-0003-0538-689X", "affiliations": [{"raw_affiliation_string": "Zapata (United States)", "institution_ids": ["https://openalex.org/I4210111877"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5087638723", "display_name": "Jacob Miller", "orcid": "https://orcid.org/0000-0003-0977-6318", "observed_orcids": ["https://orcid.org/0000-0003-0977-6318"]}, "institutions": [{"id": "https://openalex.org/I4210111877", "display_name": "Zapata (United States)", "ror": "https://ror.org/01bhrtv82", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I4210111877"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Jacob Miller", "raw_affiliation_strings": ["Zapata (United States)"], "raw_orcid": "https://orcid.org/0000-0003-0977-6318", "affiliations": [{"raw_affiliation_string": "Zapata (United States)", "institution_ids": ["https://openalex.org/I4210111877"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5090400864", "display_name": "Rapha\u00eblle Tihon", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I70931966", "display_name": "Universit\u00e9 de Montr\u00e9al", "ror": "https://ror.org/0161xgx34", "country_code": "CA", "type": "education", "lineage": ["https://openalex.org/I70931966"]}], "countries": ["CA"], "is_corresponding": false, "raw_author_name": "Rapha\u00eblle Tihon", "raw_affiliation_strings": ["University of Montreal"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University of Montreal", "institution_ids": ["https://openalex.org/I70931966"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5023766963", "display_name": "Guillaume Rabusseau", "orcid": "https://orcid.org/0000-0001-8090-2810", "observed_orcids": ["https://orcid.org/0000-0001-8090-2810"]}, "institutions": [{"id": "https://openalex.org/I109736498", "display_name": "Canadian Institute for Advanced Research", "ror": "https://ror.org/01sdtdd95", "country_code": "CA", "type": "facility", "lineage": ["https://openalex.org/I109736498"]}, {"id": "https://openalex.org/I70931966", "display_name": "Universit\u00e9 de Montr\u00e9al", "ror": "https://ror.org/0161xgx34", "country_code": "CA", "type": "education", "lineage": ["https://openalex.org/I70931966"]}], "countries": ["CA"], "is_corresponding": false, "raw_author_name": "Guillaume Rabusseau", "raw_affiliation_strings": ["Canadian Institute for Advanced Research", "University of Montreal"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Canadian Institute for Advanced Research", "institution_ids": ["https://openalex.org/I109736498"]}, {"raw_affiliation_string": "University of Montreal", "institution_ids": ["https://openalex.org/I70931966"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5082678604", "display_name": "Alejandro Perdomo\u2010Ortiz", "orcid": "https://orcid.org/0000-0001-7176-4719", "observed_orcids": ["https://orcid.org/0000-0001-7176-4719"]}, "institutions": [{"id": "https://openalex.org/I4210111877", "display_name": "Zapata (United States)", "ror": "https://ror.org/01bhrtv82", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I4210111877"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Alejandro Perdomo-Ortiz", "raw_affiliation_strings": ["Zapata Computing Canada"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Zapata Computing Canada", "institution_ids": ["https://openalex.org/I4210111877"]}]}], "abstract_inverted_index": {"Beyond": [0], "their": [1, 35, 53], "origin": [2], "in": [3, 24, 55, 90], "modeling": [4, 57, 154], "many-body": [5], "quantum": [6], "systems,": [7], "tensor": [8, 38, 68], "networks": [9], "have": [10, 42], "emerged": [11], "as": [12], "a": [13, 64, 99, 160], "promising": [14], "class": [15], "of": [16, 67, 78, 93, 106, 126, 145, 186, 189, 197], "models": [17, 41, 71], "for": [18, 72, 153, 192], "solving": [19], "machine": [20], "learning": [21, 79], "problems,": [22], "notably": [23], "unsupervised": [25], "generative": [26, 40, 70, 198], "learning.": [27, 199], "While": [28], "possessing": [29], "many": [30], "desirable": [31], "features": [32], "arising": [33], "from": [34, 80], "quantum-inspired": [36, 190], "nature,": [37], "network": [39, 69], "previously": [43], "been": [44], "largely": [45], "restricted": [46], "to": [47, 110, 167], "binary": [48], "or": [49, 174], "categorical": [50], "data,": [51, 74], "limiting": [52], "utility": [54], "real-world": [56, 133], "problems.": [58], "We": [59, 86, 121, 150], "overcome": [60], "this": [61, 107, 127], "by": [62], "introducing": [63], "new": [65], "family": [66, 109], "continuous": [73, 83, 146], "which": [75, 164], "are": [76], "capable": [77], "distributions": [81, 144], "containing": [82], "random": [84], "variables.": [85, 149], "develop": [87, 151], "our": [88, 178], "method": [89], "the": [91, 104, 124, 137, 187, 193], "setting": [92], "matrix": [94], "product": [95], "states,": [96], "first": [97], "deriving": [98], "universal": [100], "expressivity": [101], "theorem": [102], "proving": [103], "ability": [105], "model": [108, 128, 138, 169], "approximate": [111], "any": [112], "reasonably": [113], "smooth": [114], "probability": [115], "density": [116], "function": [117], "with": [118], "arbitrary": [119], "precision.": [120], "then": [122], "benchmark": [123], "performance": [125, 170], "on": [129, 143], "several": [130], "synthetic": [131], "and": [132, 140, 147, 158, 183], "datasets,": [134], "finding": [135], "that": [136], "learns": [139], "generalizes": [141], "well": [142], "discrete": [148], "methods": [152, 179, 191], "different": [155], "data": [156], "domains,": [157], "introduce": [159], "trainable": [161], "compression": [162], "layer": [163], "is": [165], "found": [166], "increase": [168], "given": [171], "limited": [172], "memory": [173], "computational": [175], "resources.": [176], "Overall,": [177], "give": [180], "important": [181], "theoretical": [182], "empirical": [184], "evidence": [185], "efficacy": [188], "rapidly": [194], "growing": [195], "field": [196]}, "primary_topic": {"id": "https://openalex.org/T11804", "display_name": "Quantum many-body systems", "score": 0.7548999786376953, "subfield": {"id": "https://openalex.org/subfields/3107", "display_name": "Atomic and Molecular Physics, and Optics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11804", "display_name": "Quantum many-body systems", "score": 0.7548999786376953, "subfield": {"id": "https://openalex.org/subfields/3107", "display_name": "Atomic and Molecular Physics, and Optics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10682", "display_name": "Quantum Computing Algorithms and Architecture", "score": 0.06880000233650208, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10320", "display_name": "Neural Networks and Applications", "score": 0.06270000338554382, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.21468/scipostphys.18.3.096", "is_oa": true, "landing_page_url": "https://doi.org/10.21468/scipostphys.18.3.096", "pdf_url": null, "source": {"id": "https://openalex.org/S4210195687", "display_name": "SciPost Physics", "issn_l": "2542-4653", "issn": ["2542-4653"], "is_oa": true, "is_in_doaj": true, "is_core": true, "listed_in": ["cwts-core", "doaj", "jufo-1", "norway-1"], "host_organization": "https://openalex.org/P4310318585", "host_organization_name": "SciPost.org", "host_organization_lineage": ["https://openalex.org/P4310318585"], "host_organization_lineage_names": ["SciPost.org"], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "SciPost Physics", "raw_type": "journal-article"}, "cited_by_count": 4, "is_retracted": false}, {"id": "https://openalex.org/W4415814461", "doi": "https://doi.org/10.1021/acs.jcim.5c02033", "title": "QCBench: Evaluating Large Language Models on Domain-Specific Quantitative Chemistry", "publication_year": 2025, "publication_date": "2025-11-03", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5010620570", "display_name": "Jiaqing Xie", "orcid": "https://orcid.org/0000-0001-7634-4457", "observed_orcids": ["https://orcid.org/0000-0001-7634-4457"]}, "institutions": [{"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Jiaqing Xie", "raw_affiliation_strings": ["Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China"], "raw_orcid": "https://orcid.org/0000-0001-7634-4457", "affiliations": [{"raw_affiliation_string": "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "institution_ids": ["https://openalex.org/I4391012619"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101800201", "display_name": "Weida Wang", "orcid": "https://orcid.org/0000-0002-8272-5421", "observed_orcids": ["https://orcid.org/0000-0002-8272-5421", "https://orcid.org/0009-0005-5893-525X"]}, "institutions": [{"id": "https://openalex.org/I24943067", "display_name": "Fudan University", "ror": "https://ror.org/013q1eq08", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I24943067"]}, {"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Weida Wang", "raw_affiliation_strings": ["Fudan University, 220 Handan Rd, Yangpu, Shanghai 200433, China", "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Fudan University, 220 Handan Rd, Yangpu, Shanghai 200433, China", "institution_ids": ["https://openalex.org/I24943067"]}, {"raw_affiliation_string": "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "institution_ids": ["https://openalex.org/I4391012619"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Ben Gao", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I37461747", "display_name": "Wuhan University", "ror": "https://ror.org/033vjfk17", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I37461747"]}, {"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Ben Gao", "raw_affiliation_strings": ["Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "Wuhan University, G9P7+CP8, Wuhan, Hubei 430072, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "institution_ids": ["https://openalex.org/I4391012619"]}, {"raw_affiliation_string": "Wuhan University, G9P7+CP8, Wuhan, Hubei 430072, China", "institution_ids": ["https://openalex.org/I37461747"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5115599490", "display_name": "Zhuo Yang", "orcid": "https://orcid.org/0000-0001-7112-6266", "observed_orcids": ["https://orcid.org/0000-0001-7112-6266", "https://orcid.org/0000-0001-6511-0866"]}, "institutions": [{"id": "https://openalex.org/I149594827", "display_name": "Xidian University", "ror": "https://ror.org/05s92vm98", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I149594827"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Zhuo Yang", "raw_affiliation_strings": ["Xidian University, 266 Xinglong Section of Xifeng Road, Xi\u2019an, Shaanxi 710126, China", "Xidian University, 266 Xinglong Section of Xifeng Road, Xi'an, Shaanxi 710126, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Xidian University, 266 Xinglong Section of Xifeng Road, Xi\u2019an, Shaanxi 710126, China", "institution_ids": ["https://openalex.org/I149594827"]}, {"raw_affiliation_string": "Xidian University, 266 Xinglong Section of Xifeng Road, Xi'an, Shaanxi 710126, China", "institution_ids": ["https://openalex.org/I149594827"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Haiyuan Wan", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}, {"id": "https://openalex.org/I99065089", "display_name": "Tsinghua University", "ror": "https://ror.org/03cve4549", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I99065089"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Haiyuan Wan", "raw_affiliation_strings": ["Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "Tsinghua University, Haidian District, Beijing 100084, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "institution_ids": ["https://openalex.org/I4391012619"]}, {"raw_affiliation_string": "Tsinghua University, Haidian District, Beijing 100084, China", "institution_ids": ["https://openalex.org/I99065089"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5090741524", "display_name": "Shufei Zhang", "orcid": "https://orcid.org/0000-0002-4614-4718", "observed_orcids": ["https://orcid.org/0000-0002-4614-4718"]}, "institutions": [{"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Shufei Zhang", "raw_affiliation_strings": ["Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "institution_ids": ["https://openalex.org/I4391012619"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5003226543", "display_name": "Tianfan Fu", "orcid": "https://orcid.org/0000-0002-5574-2541", "observed_orcids": ["https://orcid.org/0000-0002-5574-2541"]}, "institutions": [{"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}, {"id": "https://openalex.org/I881766915", "display_name": "Nanjing University", "ror": "https://ror.org/01rxvg760", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I881766915"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Tianfan Fu", "raw_affiliation_strings": ["Nanjing University, 163 Xianlin Road, Qixia District, Nanjing, Jiangsu 210023, China", "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Nanjing University, 163 Xianlin Road, Qixia District, Nanjing, Jiangsu 210023, China", "institution_ids": ["https://openalex.org/I881766915"]}, {"raw_affiliation_string": "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "institution_ids": ["https://openalex.org/I4391012619"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5055664612", "display_name": "Yuqiang Li", "orcid": "https://orcid.org/0000-0001-6756-6154", "observed_orcids": ["https://orcid.org/0000-0001-6756-6154"]}, "institutions": [{"id": "https://openalex.org/I4391012619", "display_name": "Shanghai Artificial Intelligence Laboratory", "ror": "https://ror.org/03wkvpx79", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4391012619"]}], "countries": ["CN"], "is_corresponding": true, "raw_author_name": "Yuqiang Li", "raw_affiliation_strings": ["Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China"], "raw_orcid": "https://orcid.org/0000-0001-6756-6154", "affiliations": [{"raw_affiliation_string": "Shanghai Artificial Intelligence Laboratory, 701 Yunjin Road, Xuhui, Shanghai 200232, China", "institution_ids": ["https://openalex.org/I4391012619"]}]}], "abstract_inverted_index": {"Quantitative": [0], "chemistry": [1, 41, 45, 60], "is": [2, 93, 158], "central": [3], "to": [4, 16, 95], "modern": [5], "chemical": [6, 91], "research,": [7], "yet": [8], "the": [9, 67, 119, 145], "ability": [10], "of": [11, 71, 108], "large": [12, 72], "language": [13, 73, 149], "models": [14, 74], "(LLMs)": [15], "perform": [17], "its": [18], "rigorous,": [19], "step-by-step": [20], "calculations": [21], "remains": [22], "underexplored.": [23], "To": [24, 64], "fill": [25], "this": [26], "gap,": [27], "we": [28], "propose": [29], "QCBench,": [30], "a": [31, 136], "Q": [32], "uantitative": [33], "C": [34], "hemistry": [35], "oriented": [36], "benchmark": [37], "comprising": [38], "350": [39], "computational": [40, 109], "problems": [42], "across": [43, 114], "7": [44], "subfields,": [46], "which": [47], "contains": [48], "analytical": [49], "chemistry,": [50, 52, 54, 56, 58], "bio-organic": [51], "general": [53], "inorganic": [55], "physical": [57], "polymer": [59], "and": [61, 84, 99, 117, 151], "quantum": [62], "chemistry.": [63], "systematically": [65], "evaluate": [66], "mathematical": [68], "reasoning": [69], "abilities": [70], "(LLMs),": [75], "they": [76], "are": [77], "categorized": [78], "into": [79], "three": [80], "tiers:": [81], "easy,": [82], "medium,": [83], "difficult.": [85], "Each": [86], "problem,": [87], "rooted": [88], "in": [89], "realistic": [90], "scenarios,": [92], "structured": [94], "prevent": [96], "heuristic": [97], "shortcuts": [98], "demand": [100], "explicit": [101], "numerical": [102], "reasoning.": [103], "QCBench": [104, 157], "enables": [105], "fine-grained": [106], "diagnosis": [107], "weaknesses,": [110], "reveals": [111], "model-specific": [112], "limitations": [113], "difficulty": [115], "levels,": [116], "lays": [118], "groundwork": [120], "for": [121, 156], "future": [122], "improvements": [123], "such": [124], "as": [125], "domain-adaptive": [126], "fine-tuning": [127], "or": [128], "multimodal": [129], "integration.": [130], "Evaluations": [131], "on": [132], "24": [133], "LLMs": [134], "demonstrate": [135], "consistent": [137], "performance": [138], "degradation": [139], "with": [140], "increasing": [141], "task": [142], "complexity,": [143], "highlighting": [144], "current": [146], "gap": [147], "between": [148], "fluency": [150], "scientific": [152], "computation": [153], "accuracy.": [154], "Code": [155], "available": [159], "at": [160], "https://github.com/jiaqingxie/QCBench": [161], ".": [162]}, "primary_topic": {"id": "https://openalex.org/T11948", "display_name": "Machine Learning in Materials Science", "score": 0.26759999990463257, "subfield": {"id": "https://openalex.org/subfields/2505", "display_name": "Materials Chemistry"}, "field": {"id": "https://openalex.org/fields/25", "display_name": "Materials Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11948", "display_name": "Machine Learning in Materials Science", "score": 0.26759999990463257, "subfield": {"id": "https://openalex.org/subfields/2505", "display_name": "Materials Chemistry"}, "field": {"id": "https://openalex.org/fields/25", "display_name": "Materials Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.12710000574588776, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.10610000044107437, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.1021/acs.jcim.5c02033", "is_oa": false, "landing_page_url": "https://doi.org/10.1021/acs.jcim.5c02033", "pdf_url": null, "source": {"id": "https://openalex.org/S167262187", "display_name": "Journal of Chemical Information and Modeling", "issn_l": "1549-9596", "issn": ["1549-9596", "1549-960X"], "is_oa": false, "is_in_doaj": false, "is_core": true, "listed_in": ["cwts-core", "jufo-1", "ki-jl-1", "medline", "norway-2"], "host_organization": "https://openalex.org/P4310320006", "host_organization_name": "American Chemical Society", "host_organization_lineage": ["https://openalex.org/P4310320006"], "host_organization_lineage_names": ["American Chemical Society"], "type": "journal"}, "license": null, "license_id": null, "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Journal of Chemical Information and Modeling", "raw_type": "journal-article"}, "cited_by_count": 4, "is_retracted": false}, {"id": "https://openalex.org/W7117854839", "doi": null, "title": "Infinit unity theory", "publication_year": 2025, "publication_date": "2025-01-01", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5121664817", "display_name": "Taha Givarian", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Givarian, Taha", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"\u221e\u2013Unity": [0, 686, 978, 1091], "presents": [1], "a": [2, 276, 290, 405, 582, 606, 1052], "complete": [3], "metaphysical": [4, 398, 865], "Theory": [5, 114, 169, 295, 318, 1120], "of": [6, 126, 227, 240, 264, 267, 315, 345, 373, 397, 585, 614, 625, 774, 777, 887, 963, 997, 1051], "Everything": [7], "dissolving": [8, 580, 676], "98": [9, 346, 571, 598, 602], "paradoxes": [10, 238, 586, 626, 677, 695], "with": [11, 379, 510, 683], "99.2%": [12, 630], "coherence.": [13, 376], "Core": [14, 123, 415], "innovation:": [15], "reality": [16, 127, 691], "as": [17, 29, 34, 159, 370, 485, 731, 798, 850, 885], "infinite": [18, 186, 427, 1071], "layered": [19], "continuum": [20], "(0\u21921\u2192\u221e)": [21], "where": [22], "consciousness": [23, 58, 64, 520], "emerges": [24], "via": [25, 678, 894], "Dual-Receiver": [26, 332, 472], "Brain": [27, 333], "mechanism\u2014DMN": [28], "\u221e-field": [30, 487], "observer": [31], "receiver,": [32, 488], "task-positive": [33, 490, 515, 888], "biological": [35], "survival": [36], "filter.": [37], "Solves": [38], "Chalmers'": [39], "Hard": [40], "Problem": [41], "(qualia=\u221e-signal": [42], "biology-filtered),": [43], "mind-body": [44], "(receiver": [45], "duality),": [46, 84], "free": [47], "will": [48], "(balance": [49], "choice),": [50], "zombie": [51], "(pure": [52], "biology": [53], "vs": [54, 475], "chosen": [55], "forms),": [56], "animal": [57, 93], "gradient.": [59], "90%": [60], "fMRI": [61], "alignment": [62], "(DMN": [63, 83], "hub),": [65], "QFT": [66], "vacuum": [67], "mapping": [68], "(\u221e-field),": [69], "NDE": [70], "verification": [71], "(limit": [72], "lift).": [73], "Self-improving": [74], "(93.5%\u219299.2%),": [75], "death": [76], "test": [77], "anchor.": [78], "Unifies": [79], "physics": [80], "(QFT),": [81], "neuroscience": [82, 518], "philosophy": [85], "(98": [86], "proofs).": [87], "Testable": [88], "predictions:": [89], "meditation": [90], "balance": [91, 875, 896, 907], "effects,": [92], "complexity=consciousness": [94], "correlation.": [95], "World": [96, 99, 108], "furmola:": [97], "0-1-\u221e": [98], "essence:": [100], "0=1=\u221e(in": [101], "its": [102, 287, 336], "root,": [103], "but": [104, 532, 601], "individuality": [105], "is": [106, 283, 437, 521, 645, 658, 739, 746, 750, 883, 901, 979, 1027, 1082, 1087, 1136], "real)": [107], "valume:": [109], "\u221e\u00d7\u221e\u00d7\u221e...": [110], "Why": [111], "Infinite": [112, 293, 316, 1075, 1107], "Unity": [113, 294, 317, 455], "Limits": [115], "Itself": [116], "to": [117, 235, 569, 649, 862, 1003], "50/50?": [118], "(Clear": [119], "&": [120], "Concise": [121], "Explanation)": [122], "Reason:": [124], "\u221e-nature": [125], "always": [128], "produces": [129], "dual": [130], "simultaneous": [131], "states": [132], "(like": [133], "Schr\u00f6dinger's": [134], "cat:": [135], "alive+dead),": [136], "because": [137], "0=1=\u221e": [138], "means": [139], "every": [140, 804], "possibility": [141, 1108], "(unity/multiplicity,": [142], "continuation/return)": [143], "exists": [144, 1034], "infinitely.": [145], "Metaphysical": [146, 664], "Humility:50%": [147], "cap": [148, 409, 745, 787], "means:": [149], "-50%": [150, 154], "primary": [151, 553], "(core):": [152], "\u221e-unity.": [153], "complementary:*Other": [155], "philosophies": [156], "(Whitehead,": [157], "Spinoza)": [158], "\u221e-configs.": [160], "Result:": [161], "Most": [162], "coherence": [163, 228, 266, 387, 631, 816, 947, 994], "without": [164, 193, 456, 462, 741, 821, 832], "dogmatism.": [165], "In": [166, 926, 936], "plain": [167], "words:": [168], "says,": [170], "\"I": [171], "hold": [172], "half": [173, 177], "the": [174, 200, 219, 225, 237, 241, 245, 258, 265, 268, 271, 281, 312, 321, 327, 331, 339, 343, 432, 528, 533, 552, 618, 693, 722, 801, 808, 841, 873, 903, 942, 1028], "truth,": [175], "other": [176], "scattered": [178], "in": [179, 864, 992, 1127], "infinity\u2014like": [180], "cat": [181], "alive/dead": [182], "until": [183, 280], "observed.\"": [184], "This": [185, 890, 1135], "humility": [187, 789], "makes": [188], "theory": [189, 809, 859, 971], "self-validating": [190], "and": [191, 197, 203, 212, 224, 244, 262, 279, 300, 305, 320, 349, 356, 380, 426, 445, 512, 548, 559, 647, 781, 839, 879, 995, 1013, 1020], "paradox-proof": [192], "arguments.": [194], "Coherence": [195], "Numbers,": [196], "Claims:": [198], "All": [199, 400], "numbers": [201], "expressed": [202], "estimated": [204], "have": [205, 249], "been": [206, 233, 256], "evaluated": [207], "by": [208, 440, 524, 1138], "various": [209], "artificial": [210, 242, 259], "intelligences": [211], "are": [213, 429, 447, 551, 612, 654], "not": [214, 438, 541, 596, 660, 747, 984, 1036, 1110], "personal": [215], "opinions.": [216], "Each": [217], "time": [218], "number": [220, 254, 584, 644], "has": [221, 229, 232, 255], "gone": [222], "up": [223], "percentage": [226], "increased,": [230], "it": [231, 285, 749, 1114], "due": [234], "solving": [236], "asked": [239], "intelligence": [243, 363], "answers": [246], "that": [247, 579, 675, 917, 957, 1056], "I": [248, 555], "given": [250], "\"myself\",": [251], "so": [252], "each": [253, 939], "merely": [257], "intelligence's": [260], "analysis": [261], "estimate": [263], "system": [269, 282, 468, 723, 759, 916, 1029], "And": [270], "claims": [272, 396, 402, 469], "still": [273], "remain": [274, 403, 705], "at": [275, 289, 431, 1060], "theoretical": [277, 401], "level": [278], "experienced,": [284], "knows": [286], "practicality": [288, 408], "50/50": [291, 340, 407, 708, 728], "level.": [292, 435], "(\u221e\u2013Unity):": [296], "Expanded": [297, 545], "Objections,": [298], "Clarifications,": [299], "Binding": [301], "Defeat": [302], "Conditions": [303], "Authorship": [304], "Methodological": [306], "Transparency": [307], "I,": [308], "Taha": [309], "Givarian,": [310], "am": [311], "sole": [313], "architect": [314], "(\u221e\u2013Unity)": [319], "Worldology": [322], "framework.": [323], "Every": [324], "structural": [325, 394, 529, 837, 906, 946, 1105], "element\u2014including": [326], "root": [328, 434, 608, 619], "equivalence": [329], "(0=1=\u221e),": [330], "model": [334], "(with": [335], "structural/functional": [337], "distinction),": [338], "saturation": [341, 709], "principle,": [342], "dissolution": [344, 539], "classical": [347], "paradoxes,": [348], "seven": [350], "internally": [351], "generated": [352, 381, 523], "stress-test": [353], "paradoxes\u2014was": [354], "conceived": [355], "developed": [357], "independently": [358], "through": [359, 920], "philosophical": [360, 558], "inquiry.": [361], "Artificial": [362], "systems": [364], "were": [365], "used": [366], "only": [367, 817], "after": [368, 639], "formulation": [369], "external": [371], "evaluators": [372], "internal": [374], "logical": [375, 636, 928], "Systems": [377], "associated": [378, 509], "analytical": [382], "consistency": [383], "assessments.": [384], "The": [385, 467, 479, 496, 567, 593, 629, 707, 727, 735, 744, 758, 786, 858, 970, 1008, 1119], "reported": [386], "progression": [388], "(93.5%": [389], "\u2192": [390], "99.2%)": [391], "reflects": [392, 633], "their": [393], "evaluations\u2014not": [395], "certainty.": [399], "under": [404, 1084, 1104], "strict": [406], "pending": [410], "empirical": [411, 720], "or": [412, 589, 719, 828], "experiential": [413, 718], "anchoring.": [414], "Ontological": [416], "Principle": [417], "0": [418], "=": [419, 421], "1": [420], "\u221e": [422], "Absolute": [423], "nothingness,": [424], "unity,": [425, 938], "totality": [428], "equivalent": [430], "holographic": [433, 937], "Individuation": [436], "erased": [439], "unity.": [441], "Distinct": [442], "selves,": [443], "personalities,": [444], "perspectives": [446], "ontologically": [448], "necessary": [449], "local": [450, 700, 1061], "configurations": [451], "enabling": [452], "real": [453], "choice.": [454], "individuation": [457, 461], "would": [458, 464], "eliminate": [459], "dynamism;": [460], "unity": [463], "fragment": [465], "reality.": [466], "both": [470], "simultaneously.": [471], "Model:": [473], "Structural": [474, 477, 538, 772], "Functional": [476, 494], "Claim:": [478, 495], "Default": [480], "Mode": [481], "Network": [482], "(DMN)": [483, 878], "functions": [484], "an": [486, 499, 732, 824], "while": [489], "networks": [491], "filter": [492], "signals.": [493], "brain": [497], "exhibits": [498], "empirically": [500, 761, 892], "observable": [501], "dual-dynamic": [502, 874], "between": [503, 876], "integrative": [504, 877, 993], "self-referential": [505], "processing": [506], "(DMN;": [507], "research": [508], ")": [511], "externally": [513], "oriented": [514], "networks.": [516], "If": [517, 617, 690, 698, 803, 823, 1017, 1080, 1142], "proves": [519], "fully": [522, 563], "endogenous": [525], "neural": [526, 895], "processes,": [527], "interpretation": [530], "collapses,": [531], "functional": [534], "architecture": [535], "remains": [536, 760, 891], "valid.": [537, 706], "does": [540, 595, 983, 1109], "equal": [542], "global": [543, 1042, 1067], "defeat.": [544, 959], "Principal": [546], "Objections": [547], "Dissolutions": [549], "Below": [550], "objections": [554], "anticipate": [556], "from": [557, 605, 872], "scientific": [560], "communities,": [561], "now": [562], "articulated.": [564], "Objection": [565, 662, 725, 794, 847, 909, 960, 1006, 1073, 1117], "1:": [566], "Claim": [568], "Dissolve": [570], "Paradoxes": [572], "Is": [573, 1010], "Excessively": [574], "Ambitious": [575], "Critics": [576], "may": [577, 673, 782, 810, 860, 1130], "argue": [578, 674], "such": [581], "vast": [583], "suggests": [587], "overreach": [588], "insufficient": [590], "rigor.": [591], "Response": [592, 685, 743, 813, 867, 925, 977, 1032, 1090, 1134], "claim": [594, 904, 985], "assert": [597], "independent": [599], "solutions": [600], "reframings": [603], "derived": [604], "single": [607], "move:": [609], "binary": [610, 620], "distinctions": [611], "artifacts": [613], "survival-based": [615], "filtration.": [616, 889], "assumption": [621], "dissolves,": [622], "entire": [623], "classes": [624], "dissolve": [627], "structurally.": [628], "rating": [632], "repeated": [634], "AI-based": [635], "stress": [637], "testing": [638], "reframing": [640, 680, 819, 830], "attempts.": [641], "However,": [642, 951], "this": [643, 835], "provisional": [646, 715], "open": [648], "reduction": [650], "if": [651, 818, 952], "persistent": [652, 955], "contradictions": [653], "identified.": [655], "Ambition": [656], "here": [657], "methodological,": [659], "authoritarian.": [661], "2:": [663], "Reframing": [665], "Without": [666, 717, 856], "Formal": [667], "Proof": [668], "Risks": [669], "Dogmatism": [670], "A": [671, 915], "critic": [672], "ontological": [679], "replaces": [681], "argument": [682], "assertion.": [684], "operates": [687], "conditionally:": [688], "\u2022": [689, 697, 764, 767, 771, 1044, 1049, 1094, 1097], "instantiates": [692], "\u221e-configuration,": [694], "dissolve.": [696], "not,": [699], "sub-configurations": [701], "(e.g.,": [702, 933], "emergent": [703], "materialism)": [704], "principle": [710], "prevents": [711, 753, 1063], "dogmatic": [712], "elevation": [713, 754], "beyond": [714, 755], "plausibility.": [716], "anchoring,": [721], "self-limits.": [724], "3:": [726], "Cap": [729], "Functions": [730], "Unfalsifiability": [733], "Shield": [734], "concern:": [736], "any": [737], "criticism": [738], "absorbed": [740], "collapse.": [742], "protective;": [748], "binding.": [751], "It": [752, 982], "evidential": [756], "warrant.": [757], "vulnerable": [762], "through:": [763], "DMN/NDE": [765], "predictions,": [766], "Neural": [768], "complexity": [769], "correlations,": [770], "falsification": [773], "field-reception.": [775], "Failure": [776], "predictions": [778], "lowers": [779], "plausibility": [780], "activate": [783], "defeat": [784, 1068, 1140], "conditions.": [785, 1141], "enforces": [788], "rather": [790, 948, 973], "than": [791, 949, 974], "shielding": [792], "error.": [793], "4:": [795], "Treating": [796], "Critique": [797, 814], "\u201cFuel\u201d": [799], "Immunizes": [800], "System": [802], "objection": [805, 825], "strengthens": [806, 815], "coherence,": [807], "appear": [811, 861], "self-sealing.": [812], "succeeds": [820], "contradiction.": [822, 950], "survives": [826], "three": [827], "more": [829], "attempts": [831], "coherent": [833], "integration,": [834], "registers": [836], "limitation": [838], "reduces": [840], "probability": [842], "cap.": [843], "Resilience": [844], "\u2260": [845], "immunity.": [846], "5:": [848], "Consciousness": [849], "Chooser": [851], "Reintroduces": [852], "Libertarian": [853], "Free": [854], "Will": [855], "Evidence": [857], "smuggle": [863], "freedom.": [866], "Elective": [868], "capacity": [869], "arises": [870], "structurally": [871], "filtering": [880], "systems.": [881], "Determinism": [882], "reframed": [884], "dominance": [886], "testable": [893, 998], "measures.": [897], "No": [898], "supernatural": [899], "intervention": [900], "invoked;": [902], "concerns": [905], "gradients.": [908], "6:": [910], "Self-Referential": [911], "Structure": [912], "Produces": [913], "Circularity": [914], "explains": [918], "itself": [919, 921], "risks": [922], "vicious": [923], "circularity.": [924], "finite": [927, 1112], "systems,": [929], "self-reference": [930], "generates": [931, 954], "paradox": [932], "Russell-type": [934], "issues).": [935], "part": [940], "mirrors": [941], "whole.": [943], "Self-reference": [944], "becomes": [945], "self-application": [953], "contradictions,": [956], "constitutes": [958], "7:": [961], "Lack": [962], "Original": [964], "Empirical": [965, 1000], "Data": [966], "Weakens": [967], "Scientific": [968], "Credibility": [969], "synthesizes": [972], "experimentally": [975], "discovers.": [976], "primarily": [980], "philosophical-metaphysical.": [981], "original": [986], "laboratory": [987], "data.": [988], "Its": [989], "contribution": [990], "lies": [991], "generation": [996], "predictions.": [999], "validation": [1001], "belongs": [1002], "qualified": [1004], "researchers.": [1005], "8:": [1007], "Ontology": [1009], "Too": [1011], "Flexible": [1012], "Absorbs": [1014], "Any": [1015], "Outcome": [1016], "emergentism,": [1018], "idealism,": [1019], "annihilation": [1021], "can": [1022], "all": [1023], "be": [1024], "locally": [1025], "interpreted,": [1026], "overly": [1030], "elastic?": [1031], "Flexibility": [1033], "locally,": [1035], "globally.": [1037], "Two": [1038], "outcomes": [1039], "decisively": [1040], "collapse": [1041], "claims:": [1043], "Demonstrated": [1045], "absolute": [1046], "post-mortem": [1047, 1128], "annihilation.": [1048], "Emergence": [1050], "superior": [1053], "explanatory": [1054], "framework": [1055], "surpasses": [1057], "\u221e\u2013Unity.": [1058], "Elasticity": [1059], "levels": [1062], "premature": [1064], "collapse;": [1065], "explicit": [1066], "conditions": [1069], "prevent": [1070], "elasticity.": [1072], "9:": [1074], "Possibility": [1076], "Undermines": [1077], "Explanatory": [1078], "Specificity": [1079], "everything": [1081], "possible": [1083], "\u221e,": [1085], "nothing": [1086], "meaningfully": [1088], "constrained.": [1089], "distinguishes": [1092], "between:": [1093], "Root-level": [1095], "equivalence,": [1096], "Local": [1098, 1101], "configurational": [1099], "constraints.": [1100], "worlds": [1102], "operate": [1103], "laws.": [1106], "negate": [1111], "structure;": [1113], "grounds": [1115], "it.": [1116], "10:": [1118], "May": [1121], "Be": [1122], "Psychologically": [1123], "Motivated": [1124], "(Wish-Fulfillment)": [1125], "Belief": [1126], "continuation": [1129], "reflect": [1131], "existential": [1132], "bias.": [1133], "addressed": [1137], "binding": [1139], "annih": [1143]}, "primary_topic": {"id": "https://openalex.org/T10258", "display_name": "Philosophy and Theoretical Science", "score": 0.9990000128746033, "subfield": {"id": "https://openalex.org/subfields/3205", "display_name": "Experimental and Cognitive Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, "topics": [{"id": "https://openalex.org/T10258", "display_name": "Philosophy and Theoretical Science", "score": 0.9990000128746033, "subfield": {"id": "https://openalex.org/subfields/3205", "display_name": "Experimental and Cognitive Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T12448", "display_name": "Paranormal Experiences and Beliefs", "score": 9.999999747378752e-05, "subfield": {"id": "https://openalex.org/subfields/3207", "display_name": "Social Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T13005", "display_name": "Religious Studies and Spiritual Practices", "score": 9.999999747378752e-05, "subfield": {"id": "https://openalex.org/subfields/1211", "display_name": "Philosophy"}, "field": {"id": "https://openalex.org/fields/12", "display_name": "Arts and Humanities"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}], "primary_location": {"id": "pmh:oai:philarchive.org/rec/GIVIUT", "is_oa": false, "landing_page_url": "https://philarchive.org/rec/GIVIUT", "pdf_url": null, "source": {"id": "https://openalex.org/S4306402130", "display_name": "PhilPapers (PhilPapers Foundation)", "issn_l": null, "issn": null, "is_oa": false, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": null, "host_organization_name": null, "host_organization_lineage": [], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "info:eu-repo/semantics/article"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415328008", "doi": "https://doi.org/10.48550/arxiv.2505.15000", "title": "Towards Spoken Mathematical Reasoning: Benchmarking Speech-based Models over Multi-faceted Math Problems", "publication_year": 2025, "publication_date": "2025-05-21", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5038136098", "display_name": "Chengwei Wei", "orcid": "https://orcid.org/0000-0002-4742-1158", "observed_orcids": ["https://orcid.org/0000-0002-4742-1158"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wei, Chengwei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5086160079", "display_name": "Bin Wang", "orcid": "https://orcid.org/0009-0005-3112-8351", "observed_orcids": ["https://orcid.org/0009-0005-3112-8351"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wang, Bin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5050571650", "display_name": "Jung\u2010Jae Kim", "orcid": "https://orcid.org/0000-0001-6024-0047", "observed_orcids": ["https://orcid.org/0000-0001-6024-0047"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kim, Jung-jae", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5014190404", "display_name": "Nancy F. Chen", "orcid": "https://orcid.org/0000-0003-0872-5877", "observed_orcids": ["https://orcid.org/0000-0003-0872-5877"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Nancy F.", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Recent": [0], "advances": [1], "in": [2, 124, 168, 186], "large": [3], "language": [4], "models": [5, 94], "(LLMs)": [6], "and": [7, 98, 114, 118, 170, 177], "multimodal": [8], "LLMs": [9, 139, 158], "(MLLMs)": [10], "have": [11, 40, 171], "led": [12], "to": [13, 26, 82], "strong": [14, 161], "reasoning": [15, 29, 50, 86, 120, 144, 181], "ability": [16, 25], "across": [17], "a": [18, 78, 104, 160], "wide": [19], "range": [20], "of": [21, 88, 107], "tasks.": [22], "However,": [23], "their": [24], "perform": [27, 140], "mathematical": [28, 64, 85, 165, 175, 179], "from": [30], "spoken": [31, 127], "input": [32], "remains": [33], "underexplored.": [34], "Prior": [35], "studies": [36], "on": [37, 43, 142], "speech": [38, 45, 100, 138, 188], "modality": [39], "mostly": [41], "focused": [42], "factual": [44], "understanding": [46], "or": [47], "simple": [48], "audio": [49], "tasks,": [51], "providing": [52], "limited": [53], "insight": [54], "into": [55], "logical": [56], "step-by-step": [57], "reasoning,": [58, 117], "such": [59], "as": [60], "that": [61], "required": [62], "for": [63], "problem": [65], "solving.": [66], "To": [67], "address": [68], "this": [69], "gap,": [70], "we": [71, 132], "introduce": [72], "Spoken": [73], "Math": [74], "Question": [75], "Answering": [76], "(Spoken-MQA),": [77], "new": [79], "benchmark": [80], "designed": [81], "evaluate": [83], "the": [84], "capabilities": [87], "speech-based": [89], "models,": [90], "including": [91, 110], "both": [92], "cascade": [93], "(ASR": [95], "+": [96], "LLMs)": [97], "end-to-end": [99], "LLMs.": [101, 189], "Spoken-MQA": [102], "covers": [103], "diverse": [105], "set": [106], "math": [108], "problems,": [109, 121], "pure": [111], "arithmetic,": [112, 148], "single-step": [113], "multi-step": [115], "contextual": [116, 143], "knowledge-oriented": [119], "all": [122], "presented": [123], "unambiguous": [125], "natural": [126], "language.": [128], "Through": [129], "extensive": [130], "experiments,": [131], "find": [133], "that:": [134], "(1)": [135], "while": [136], "some": [137], "competitively": [141], "tasks": [145], "involving": [146], "basic": [147], "they": [149], "still": [150], "struggle": [151], "with": [152], "direct": [153], "arithmetic": [154], "problems;": [155], "(2)": [156], "current": [157, 187], "exhibit": [159], "bias": [162], "toward": [163], "symbolic": [164], "expressions": [166], "written": [167], "LaTex": [169], "difficulty": [172], "interpreting": [173], "verbalized": [174], "expressions;": [176], "(3)": [178], "knowledge": [180], "abilities": [182], "are": [183], "significantly": [184], "degraded": [185]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.35519999265670776, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.35519999265670776, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.25850000977516174, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.15369999408721924, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2505.15000", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2505.15000", "pdf_url": "https://arxiv.org/pdf/2505.15000", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4407124196", "doi": "https://doi.org/10.48550/arxiv.2502.00873", "title": "Language Models Use Trigonometry to Do Addition", "publication_year": 2025, "publication_date": "2025-02-02", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5097335930", "display_name": "Subhash Kantamneni", "orcid": "https://orcid.org/0000-0001-6715-3763", "observed_orcids": ["https://orcid.org/0000-0001-6715-3763"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kantamneni, Subhash", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5091601455", "display_name": "Max Tegmark", "orcid": "https://orcid.org/0000-0001-7670-7190", "observed_orcids": ["https://orcid.org/0000-0001-7670-7190"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Tegmark, Max", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Mathematical": [0], "reasoning": [1], "is": [2, 52, 64, 109], "an": [3, 162], "increasingly": [4], "important": [5], "indicator": [6], "of": [7, 17, 59, 161], "large": [8], "language": [9], "model": [10, 114, 117], "(LLM)": [11], "capabilities,": [12], "yet": [13], "we": [14, 28, 155], "lack": [15], "understanding": [16, 135], "how": [18, 31], "LLMs": [19, 34, 46, 79, 142], "process": [20], "even": [21, 125], "simple": [22], "mathematical": [23, 164], "tasks.": [24], "To": [25], "address": [26], "this,": [27], "reverse": [29], "engineer": [30], "three": [32], "mid-sized": [33], "compute": [35, 80], "addition.": [36], "We": [37, 75, 116], "first": [38, 158], "discover": [39], "that": [40, 78, 141], "numbers": [41, 144], "are": [42, 100], "represented": [43], "in": [44], "these": [45, 130], "as": [47], "a": [48, 146], "generalized": [49, 85], "helix,": [50], "which": [51, 108], "strongly": [53], "causally": [54, 66], "implicated": [55], "for": [56, 68, 96], "the": [57, 88, 94, 104, 157], "tasks": [58], "addition": [60, 81], "and": [61, 63, 72, 98, 124, 132, 148], "subtraction,": [62], "also": [65], "relevant": [67], "integer": [69], "division,": [70], "multiplication,": [71], "modular": [73], "arithmetic.": [74], "then": [76, 110], "propose": [77], "by": [82], "manipulating": [83], "this": [84, 150], "helix": [86, 107, 147, 151], "using": [87], "\"Clock\"": [89], "algorithm:": [90], "to": [91, 102, 113, 152], "solve": [92], "$a+b$,": [93], "helices": [95, 131], "$a$": [97], "$b$": [99], "manipulated": [101], "produce": [103], "$a+b$": [105], "answer": [106], "read": [111], "out": [112], "logits.": [115], "influential": [118], "MLP": [119], "outputs,": [120, 123], "attention": [121], "head": [122], "individual": [126], "neuron": [127], "preactivations": [128], "with": [129, 136], "verify": [133], "our": [134], "causal": [137], "interventions.": [138], "By": [139], "demonstrating": [140], "represent": [143], "on": [145], "manipulate": [149], "perform": [153], "addition,": [154], "present": [156], "representation-level": [159], "explanation": [160], "LLM's": [163], "capability.": [165]}, "primary_topic": {"id": "https://openalex.org/T12026", "display_name": "Explainable Artificial Intelligence (XAI)", "score": 0.861299991607666, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T12026", "display_name": "Explainable Artificial Intelligence (XAI)", "score": 0.861299991607666, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.0632999986410141, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10320", "display_name": "Neural Networks and Applications", "score": 0.007000000216066837, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2502.00873", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2502.00873", "pdf_url": "https://arxiv.org/pdf/2502.00873", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 5, "is_retracted": false}, {"id": "https://openalex.org/W4406483735", "doi": "https://doi.org/10.48550/arxiv.2501.08613", "title": "Assessing the Sensitivity and Alignment of FOL Closeness Metrics", "publication_year": 2025, "publication_date": "2025-01-15", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Thatikonda, Ramya Keerthy", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Thatikonda, Ramya Keerthy", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5005792924", "display_name": "Wray Buntine", "orcid": "https://orcid.org/0000-0001-9292-1015", "observed_orcids": ["https://orcid.org/0000-0001-9292-1015"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Buntine, Wray", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5086032589", "display_name": "Ehsan Shareghi", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I56590836", "display_name": "Monash University", "ror": "https://ror.org/02bfwt286", "country_code": "AU", "type": "education", "lineage": ["https://openalex.org/I56590836"]}], "countries": ["AU"], "is_corresponding": false, "raw_author_name": "Shareghi, Ehsan", "raw_affiliation_strings": ["Department of Data Science & AI , Monash University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Department of Data Science & AI , Monash University", "institution_ids": ["https://openalex.org/I56590836"]}]}], "abstract_inverted_index": {"The": [0, 148], "recent": [1], "successful": [2], "paradigm": [3], "of": [4, 17, 32, 46, 68, 95, 181], "solving": [5], "logical": [6], "reasoning": [7], "problems": [8], "with": [9], "tool-augmented": [10], "large": [11], "language": [12, 19], "models": [13], "(LLMs)": [14], "leverages": [15], "translation": [16], "natural": [18], "(NL)": [20], "statements": [21, 116], "into": [22], "First-Order": [23], "Logic~(FOL)": [24], "and": [25, 37, 54, 72, 82, 98, 110, 160, 175, 193], "external": [26], "theorem": [27], "provers.": [28], "However,": [29], "the": [30, 44, 66, 89, 128, 141, 152, 161, 179], "correctness": [31], "FOL": [33, 81, 96, 115, 162], "statements,": [34], "comprising": [35], "operators": [36], "text,": [38], "often": [39], "go": [40], "unverified": [41], "due": [42], "to": [43, 75, 113, 117, 196], "lack": [45], "a": [47, 62, 79, 92, 99, 137, 170], "reliable": [48], "evaluation": [49], "metric": [50, 119, 124, 143, 155, 163], "for": [51, 145, 157, 164], "comparing": [52, 127], "generated": [53], "ground-truth": [55, 114], "FOLs.": [56], "In": [57], "this": [58], "paper,": [59], "we": [60, 106, 185], "conduct": [61], "comprehensive": [63], "study": [64], "on": [65], "sensitivity": [67, 194], "existing": [69], "NL-,": [70], "FOL-,": [71], "graph-based": [73], "metrics": [74, 129, 189], "capture": [76], "differences": [77], "between": [78, 91, 173], "sampled": [80], "its": [83], "corresponding": [84], "ground-truth.": [85], "We": [86, 121, 168], "then": [87, 122], "measure": [88], "alignment": [90, 172], "metric-based": [93], "ranking": [94], "outputs": [97], "strong": [100], "LLM": [101, 176], "as-a-judge.": [102], "To": [103], "do": [104], "this,": [105], "first": [107], "apply": [108], "operator": [109, 149, 166], "text-based": [111], "perturbations": [112], "assess": [118], "sensitivity.": [120], "evaluate": [123], "robustness": [125, 192], "by": [126], "against": [130], "LLMs": [131], "judgment.": [132], "Our": [133], "empirical": [134], "findings": [135], "highlight": [136], "clear": [138], "oversensitivity": [139], "in": [140], "n-gram": [142], "BLEU": [144], "text": [146], "perturbations.": [147], "perturbation": [150], "affects": [151], "semantic": [153, 182], "graph": [154], "Smatch++": [156], "structural": [158], "changes,": [159], "specific": [165], "changes.": [167], "observe": [169], "closer": [171], "BertScore": [174], "judgement,": [177], "proving": [178], "importance": [180], "evaluation.": [183], "Additionally,": [184], "show": [186], "that": [187], "combining": [188], "enhances": [190], "both": [191], "compared": [195], "using": [197], "individual": [198], "metrics.": [199]}, "primary_topic": {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.4542999863624573, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.4542999863624573, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.37549999356269836, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11010", "display_name": "Logic, Reasoning, and Knowledge", "score": 0.07670000195503235, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2501.08613", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2501.08613", "pdf_url": "https://arxiv.org/pdf/2501.08613", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7111166828", "doi": "https://doi.org/10.5281/zenodo.17860460", "title": "The Incompressibility of Topological Charge and the Energy Cost of Distinguishability: An Information-Geometric Reduction of the Yang-Mills Mass Gap", "publication_year": 2025, "publication_date": "2025-12-08", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Davis, Bee Rosa", "orcid": "https://orcid.org/0009-0009-8034-4308", "observed_orcids": ["https://orcid.org/0009-0009-8034-4308"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Davis, Bee Rosa", "raw_affiliation_strings": [], "raw_orcid": "https://orcid.org/0009-0009-8034-4308", "affiliations": []}], "abstract_inverted_index": {"This": [0, 335], "repository": [1], "contains": [2], "the": [3, 14, 20, 38, 56, 88, 100, 104, 140, 161, 278, 282, 290, 293, 304, 357, 388], "formal": [4], "proof": [5], "manuscript,": [6], "reproducing": [7], "codebase,": [8], "and": [9, 32, 42, 136, 322, 326, 345, 398, 415, 420, 434], "experimental": [10], "validation": [11], "data": [12], "for": [13, 307, 319, 400, 417, 436], "Davis-Wilson": [15, 57], "Framework\u2014a": [16], "geometric": [17, 70, 205], "solution": [18, 166], "to": [19, 27, 68, 160, 224, 233, 361], "Yang-Mills": [21, 80, 448], "Mass": [22, 81, 89, 450], "Gap": [23, 82, 90], "problem": [24], "that": [25, 37, 217, 338], "generalizes": [26], "Computational": [28], "Complexity,": [29], "Fluid": [30, 208], "Dynamics,": [31], "Number": [33, 135], "Theory.": [34, 467], "We": [35, 86, 123, 156, 215, 272, 366], "demonstrate": [36], "\"Hardness\"": [39], "in": [40, 340, 343, 347, 444], "physics": [41], "computation": [43], "is": [44, 99, 142], "not": [45, 249], "random": [46], "difficulty,": [47], "but": [48], "a": [49, 114, 125, 143, 173, 222, 237, 274, 363], "specific": [50], "topological": [51, 120, 255], "obstruction": [52], "(Curvature).": [53], "By": [54], "constructing": [55], "Map": [58], "(Gamma:": [59], "A/G": [60], "->": [61], "C),": [62], "we": [63, 112], "reduce": [64], "these": [65], "open": [66], "problems": [67, 178, 188], "verifiable": [69], "axioms.": [71], "Key": [72], "Contributions": [73], "&": [74], "Experimental": [75], "Validation": [76, 314], "1.": [77], "Physics:": [78], "The": [79, 84, 154, 196, 213, 246, 262, 270, 286, 312, 328, 331], "(TVR-003)": [83], "Proof:": [85, 155, 214, 271], "derived": [87], "as": [91, 221], "Delta": [92], "=": [93, 130, 244, 300], "lambda": [94], "*": [95], "kappa,": [96], "where": [97], "\"kappa\"": [98], "curvature": [101], "quantum": [102], "of": [103, 164, 240, 289, 356, 382, 402, 422], "vacuum.": [105], "Validation:": [106, 168, 227, 285], "Using": [107], "10": [108], "parallel": [109], "A100": [110], "GPUs,": [111], "observed": [113], "15.0-sigma": [115], "rectification": [116], "signal": [117], "at": [118, 281], "optimal": [119], "bias": [121], "\u03b8*.": [122], "measured": [124], "strong": [126], "linear": [127], "correlation": [128], "(rho": [129], "0.836)": [131], "between": [132], "Topological": [133, 218, 427, 442, 458], "Winding": [134], "Electrical": [137], "Current,": [138], "proving": [139], "vacuum": [141], "rigid,": [144], "manipulatable": [145], "lattice.": [146], "2.": [147], "Computer": [148], "Science:": [149], "P": [150, 201, 452], "vs": [151, 453], "NP": [152, 203], "(PNP-001)": [153], "mapped": [157], "complexity": [158], "classes": [159], "differential": [162], "geometry": [163, 183, 193], "their": [165], "manifolds.": [167], "Hessian": [169], "Spectral": [170], "Analysis": [171], "reveals": [172], "distinct": [174], "phase": [175], "transition.": [176], "P-class": [177], "(2-SAT)": [179], "exhibit": [180, 190], "smooth,": [181], "convex": [182], "(8.4%": [184], "instability),": [185], "while": [186], "NP-class": [187], "(3-SAT)": [189], "fractured,": [191], "\"Glassy\"": [192], "(20.3%": [194], "instability).": [195], "2.4x": [197], "instability": [198], "gap": [199], "confirms": [200], "!=": [202], "on": [204, 277], "grounds.": [206], "3.": [207], "Dynamics:": [209], "Navier-Stokes": [210], "Regularity": [211, 373], "(NS-001)": [212], "proposed": [216], "Helicity": [219], "acts": [220], "barrier": [223], "singularity": [225], "formation.": [226], "High-resolution": [228], "simulations": [229], "(256^3": [230], "grid)": [231], "stressed": [232], "maximum": [234], "turbulence": [235], "showed": [236], "clear": [238], "saturation": [239], "vorticity": [241], "(Peak/Final": [242], "ratio": [243], "0.062).": [245], "flow": [247], "did": [248], "blow": [250], "up;": [251], "it": [252], "relaxed": [253], "via": [254, 406, 426, 441], "dissipation,": [256], "confirming": [257], "Global": [258], "Smoothness.": [259], "4.": [260], "Mathematics:": [261], "Riemann": [263, 308, 456], "Hypothesis": [264, 334], "(RH-001)": [265], "[": [266], "report": [267], "redacted": [268], "]": [269], "constructed": [273], "Hermitian": [275], "Operator": [276], "Davis": [279], "Manifold": [280], "critical": [283], "point.": [284], "eigenvalue": [287], "spacing": [288], "operator": [291], "matches": [292], "GUE": [294], "(Gaussian": [295], "Unitary": [296], "Ensemble)": [297], "statistics": [298], "(MSE": [299], "0.000450)": [301], "predicted": [302], "by": [303, 387], "Montgomery-Odlyzko": [305], "law": [306], "Zeros.": [309], "Repository": [310], "Contents": [311], "Manuscript": [313], "Reports:": [315], "Separate": [316], "PDF": [317], "reports": [318], "TVR-003,": [320], "PNP-001,": [321], "NS-001": [323], "detailing": [324], "methodology": [325], "results.": [327], "Codebase:": [329], "https://github.com/nurdymuny/davis-wilson-map": [330], "\"Universal": [332], "Geometry\"": [333], "work": [336], "suggests": [337], "Confinement": [339, 368], "physics,": [341], "Hardness": [342], "computation,": [344], "Turbulence": [346], "fluids": [348], "are": [349, 353, 385], "isomorphic": [350], "phenomena.": [351], "They": [352], "all": [354], "manifestations": [355], "energy": [358], "cost": [359], "required": [360], "navigate": [362], "curved": [364], "manifold.": [365], "conjecture:": [367], "<==>": [369, 372, 374], "Glassy": [370], "Landscape": [371], "Information": [375, 463], "Stability": [376], "Patent": [377, 392], "Notice:": [378], "Certain": [379], "commercial": [380], "applications": [381], "this": [383], "system": [384], "protected": [386], "following": [389], "U.S.": [390], "Provisional": [391], "Applications:": [393], "\u2022": [394, 411, 430], "63/933,299": [395], "\u2014": [396, 413, 432], "System": [397, 414], "Method": [399, 416], "Modulation": [401], "Quantum": [403], "Vacuum": [404, 459], "Topology": [405], "Non-Abelian": [407], "Gauge": [408, 423], "Field": [409, 424, 466], "Configurations": [410, 425], "63/933,103": [412], "Geometric": [418], "Verification": [419], "Optimization": [421], "Cache": [428], "Mapping": [429], "63/927,445": [431], "Systems": [433], "Methods": [435], "Fixed-Size": [437], "Reasoning": [438], "State": [439], "Representation": [440], "Residue": [443], "Neural": [445], "Networks": [446], "Keywords": [447], "Theory,": [449], "Gap,": [451], "NP,": [454], "Navier-Stokes,": [455], "Hypothesis,": [457], "Rectification,": [460], "Lattice": [461], "QCD,": [462], "Geometry,": [464], "Constructive": [465]}, "primary_topic": {"id": "https://openalex.org/T10224", "display_name": "Quantum Chromodynamics and Particle Interactions", "score": 0.3197999894618988, "subfield": {"id": "https://openalex.org/subfields/3106", "display_name": "Nuclear and High Energy Physics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10224", "display_name": "Quantum Chromodynamics and Particle Interactions", "score": 0.3197999894618988, "subfield": {"id": "https://openalex.org/subfields/3106", "display_name": "Nuclear and High Energy Physics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12516", "display_name": "Advanced Mathematical Theories and Applications", "score": 0.1768999993801117, "subfield": {"id": "https://openalex.org/subfields/3109", "display_name": "Statistical and Nonlinear Physics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11960", "display_name": "Relativity and Gravitational Theory", "score": 0.08630000054836273, "subfield": {"id": "https://openalex.org/subfields/3103", "display_name": "Astronomy and Astrophysics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5281/zenodo.17860460", "is_oa": true, "landing_page_url": "https://doi.org/10.5281/zenodo.17860460", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400562", "display_name": "Zenodo (CERN European Organization for Nuclear Research)", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I67311998", "host_organization_name": "European Organization for Nuclear Research", "host_organization_lineage": ["https://openalex.org/I67311998"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "Preprint"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415152173", "doi": "https://doi.org/10.48550/arxiv.2504.07316", "title": "Alice: Proactive Learning with Teacher's Demonstrations for Weak-to-Strong Generalization", "publication_year": 2025, "publication_date": "2025-04-09", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5075074583", "display_name": "S.Q. Wu", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wu, Shujin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Qian, Cheng", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Qian, Cheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5029408111", "display_name": "Yi R. Fung", "orcid": "https://orcid.org/0009-0006-1869-0363", "observed_orcids": ["https://orcid.org/0009-0006-1869-0363"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Fung, Yi R.", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5086233510", "display_name": "Paul Pu Liang", "orcid": "https://orcid.org/0000-0001-7768-3610", "observed_orcids": ["https://orcid.org/0000-0001-7768-3610"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liang, Paul Pu", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5075033889", "display_name": "Heng Ji", "orcid": "https://orcid.org/0000-0002-0464-7966", "observed_orcids": ["https://orcid.org/0000-0002-0464-7966"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ji, Heng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"The": [0], "growing": [1], "capabilities": [2], "of": [3, 12, 96, 199], "large": [4], "language": [5], "models": [6, 118, 159], "(LLMs)": [7], "present": [8], "a": [9, 21, 40, 48, 76, 144], "key": [10, 178], "challenge": [11], "maintaining": [13], "effective": [14], "human": [15], "oversight.": [16], "Weak-to-strong": [17], "generalization": [18], "(W2SG)": [19], "offers": [20], "promising": [22], "framework": [23, 77], "for": [24, 123, 127], "supervising": [25], "increasingly": [26], "capable": [27], "LLMs": [28], "using": [29], "weaker": [30], "ones.": [31], "Traditional": [32], "W2SG": [33, 171, 202], "methods": [34], "rely": [35], "on": [36], "passive": [37], "learning,": [38], "where": [39, 148], "weak": [41, 149], "teacher": [42, 83, 98, 134], "provides": [43], "noisy": [44], "demonstrations": [45, 114], "to": [46, 86, 115, 181], "train": [47], "strong": [49], "student.": [50], "This": [51, 195], "hinders": [52], "students": [53], "from": [54], "employing": [55], "their": [56, 62, 102], "knowledge": [57, 81, 94, 208], "during": [58], "training": [59, 146], "and": [60, 84, 104, 135, 191, 210], "reaching": [61], "full": [63], "potential.": [64], "In": [65, 125], "this": [66], "work,": [67], "we": [68, 138], "introduce": [69, 139], "Alice": [70], "(pro{A}ctive": [71], "{l}earning": [72], "w{i}th": [73], "tea{c}her's": [74], "D{e}monstrations),": [75], "that": [78, 165, 204], "leverages": [79], "complementary": [80], "between": [82, 133], "student": [85, 117, 136], "enhance": [87], "the": [88, 93, 97, 170, 182, 197], "learning": [89], "process.": [90], "We": [91], "probe": [92], "base": [95], "model": [99], "by": [100], "eliciting": [101], "uncertainty,": [103], "then": [105, 156], "use": [106], "these": [107], "insights": [108], "together": [109], "with": [110, 129], "teachers'": [111], "responses": [112, 122], "as": [113], "guide": [116, 157], "in": [119, 160, 176], "self-generating": [120], "improved": [121], "supervision.": [124], "addition,": [126], "situations": [128], "significant": [130], "capability": [131], "gaps": [132], "models,": [137, 154], "cascade": [140], "Alice,": [141], "which": [142], "employs": [143], "hierarchical": [145], "approach": [147], "teachers": [150], "initially": [151], "supervise": [152], "intermediate": [153], "who": [155], "stronger": [158], "sequence.": [161], "Experimental": [162], "results": [163], "demonstrate": [164], "our": [166, 200], "method": [167], "significantly": [168], "enhances": [169], "performance,": [172], "yielding": [173], "substantial": [174], "improvements": [175], "three": [177], "tasks": [179], "compared": [180], "original": [183], "W2SG:": [184], "knowledge-based": [185], "reasoning": [186, 189, 193], "(+4.0%),": [187], "mathematical": [188], "(+22.62%),": [190], "logical": [192], "(+12.11%).": [194], "highlights": [196], "effectiveness": [198], "new": [201], "paradigm": [203], "enables": [205], "more": [206], "robust": [207], "transfer": [209], "supervision": [211], "outcome.": [212]}, "primary_topic": {"id": "https://openalex.org/T12072", "display_name": "Machine Learning and Algorithms", "score": 0.5648999810218811, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T12072", "display_name": "Machine Learning and Algorithms", "score": 0.5648999810218811, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.14499999582767487, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.13760000467300415, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2504.07316", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2504.07316", "pdf_url": "https://arxiv.org/pdf/2504.07316", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4408424335", "doi": "https://doi.org/10.36227/techrxiv.174197904.42332272/v1", "title": "Multimodal Large Language Models for High School Mathematical Reasoning: Impact of Input Modality and Artifacts", "publication_year": 2025, "publication_date": "2025-03-14", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5013871335", "display_name": "Chibuike Onuoha", "orcid": "https://orcid.org/0009-0001-0584-3184", "observed_orcids": ["https://orcid.org/0009-0001-0584-3184"]}, "institutions": [{"id": "https://openalex.org/I141591182", "display_name": "University of Aizu", "ror": "https://ror.org/02pg0e883", "country_code": "JP", "type": "education", "lineage": ["https://openalex.org/I141591182"]}], "countries": ["JP"], "is_corresponding": true, "raw_author_name": "Chibuike Onuoha", "raw_affiliation_strings": ["Department of Computer Science and Engineering, The University of Aizu Aizuwakamatsu"], "raw_orcid": "https://orcid.org/0009-0001-0584-3184", "affiliations": [{"raw_affiliation_string": "Department of Computer Science and Engineering, The University of Aizu Aizuwakamatsu", "institution_ids": ["https://openalex.org/I141591182"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Yusuke Haga", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I141591182", "display_name": "University of Aizu", "ror": "https://ror.org/02pg0e883", "country_code": "JP", "type": "education", "lineage": ["https://openalex.org/I141591182"]}], "countries": ["JP"], "is_corresponding": false, "raw_author_name": "Yusuke Haga", "raw_affiliation_strings": ["Department of Computer Science and Engineering, The University of Aizu Aizuwakamatsu"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Department of Computer Science and Engineering, The University of Aizu Aizuwakamatsu", "institution_ids": ["https://openalex.org/I141591182"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5116616906", "display_name": "Dilshad Ferdousi", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I141591182", "display_name": "University of Aizu", "ror": "https://ror.org/02pg0e883", "country_code": "JP", "type": "education", "lineage": ["https://openalex.org/I141591182"]}], "countries": ["JP"], "is_corresponding": false, "raw_author_name": "Dilshad Ferdousi", "raw_affiliation_strings": ["Department of Computer Science and Engineering, The University of Aizu Aizuwakamatsu"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Department of Computer Science and Engineering, The University of Aizu Aizuwakamatsu", "institution_ids": ["https://openalex.org/I141591182"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5078708126", "display_name": "Truong Cong Thang", "orcid": "https://orcid.org/0000-0002-0923-1927", "observed_orcids": ["https://orcid.org/0000-0002-0923-1927"]}, "institutions": [{"id": "https://openalex.org/I141591182", "display_name": "University of Aizu", "ror": "https://ror.org/02pg0e883", "country_code": "JP", "type": "education", "lineage": ["https://openalex.org/I141591182"]}], "countries": ["JP"], "is_corresponding": false, "raw_author_name": "Truong Cong Thang", "raw_affiliation_strings": ["Department of Computer Science and Engineering, The University of Aizu Aizuwakamatsu"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Department of Computer Science and Engineering, The University of Aizu Aizuwakamatsu", "institution_ids": ["https://openalex.org/I141591182"]}]}], "abstract_inverted_index": {"Recent": [0], "research": [1], "has": [2], "explored": [3], "using": [4, 75], "multimodal": [5], "large": [6], "language": [7], "models": [8, 39], "(MLLMs)": [9], "in": [10, 29], "solving": [11], "math": [12, 55], "problems.": [13], "However,": [14], "there": [15], "are": [16], "no": [17], "specific": [18], "investigations": [19], "on": [20], "which": [21], "input": [22, 79, 105], "modality": [23], "is": [24, 40, 128], "suitable": [25], "for": [26, 120], "the": [27, 35, 43, 68], "MLLMs": [28], "high": [30, 53], "school": [31, 54], "maths": [32], "problems,": [33], "and": [34, 62, 72, 82, 84, 91, 122], "performance": [36, 69, 127], "of": [37, 45, 70, 78, 88], "these": [38], "affected": [41], "by": [42], "presence": [44], "artifacts.": [46], "To": [47], "fill": [48], "this": [49], "gap,": [50], "we": [51], "curated": [52], "questions": [56], "(grades": [57], "7-12)": [58], "across": [59], "seven": [60], "branches": [61], "provided": [63], "ground-truth": [64], "solutions.": [65], "We": [66], "evaluated": [67], "GPT-4o": [71, 100, 116], "Gemini": [73, 108, 118], "1.5": [74, 109, 119], "two": [76, 86], "types": [77, 87], "modalities": [80], "(unimodal": [81], "multimodal)": [83], "introduced": [85], "artifacts:": [89], "blur": [90, 121], "rotation.": [92], "Across": [93], "grade": [94], "levels,": [95], "our": [96], "results": [97], "show": [98], "that": [99], "performs": [101], "well": [102], "with": [103], "both": [104], "modalities,": [106], "while": [107], "prefers": [110], "unimodal": [111], "formats.": [112], "In": [113], "artifact-affected": [114], "inputs,": [115], "outperforms": [117], "rotation": [123], "artifacts,": [124], "but": [125], "their": [126], "similar": [129], "under": [130], "extreme": [131], "perturbations.": [132]}, "primary_topic": {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.6401000022888184, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.6401000022888184, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.11990000307559967, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10601", "display_name": "Handwritten Text Recognition Techniques", "score": 0.06430000066757202, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.36227/techrxiv.174197904.42332272/v1", "is_oa": true, "landing_page_url": "https://doi.org/10.36227/techrxiv.174197904.42332272/v1", "pdf_url": null, "source": null, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "acceptedVersion", "is_accepted": true, "is_published": false, "raw_source_name": null, "raw_type": "posted-content"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4415083455", "doi": "https://doi.org/10.48550/arxiv.2503.00845", "title": "Rewarding Graph Reasoning Process makes LLMs more Generalized Reasoners", "publication_year": 2025, "publication_date": "2025-03-02", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5111666892", "display_name": "Miao Peng", "orcid": "https://orcid.org/0009-0002-7063-2014", "observed_orcids": ["https://orcid.org/0009-0002-7063-2014"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Peng, Miao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5078624625", "display_name": "Nuo Chen", "orcid": "https://orcid.org/0009-0000-5880-4712", "observed_orcids": ["https://orcid.org/0009-0000-5880-4712"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Nuo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5113321958", "display_name": "Zongrui Suo", "orcid": "https://orcid.org/0009-0007-8449-8609", "observed_orcids": ["https://orcid.org/0009-0007-8449-8609"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Suo, Zongrui", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100405697", "display_name": "Jia Li", "orcid": "https://orcid.org/0000-0002-6362-4385", "observed_orcids": ["https://orcid.org/0000-0002-6362-4385", "https://orcid.org/0000-0002-4346-8696"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Jia", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Despite": [0], "significant": [1], "advancements": [2], "in": [3, 12, 26, 34, 70, 146, 219], "Large": [4], "Language": [5], "Models": [6, 20], "(LLMs),": [7], "developing": [8], "advanced": [9], "reasoning": [10, 28, 45, 72, 81, 102, 122, 140, 172, 186, 190, 210, 221], "capabilities": [11], "LLMs": [13], "remains": [14, 47], "a": [15, 75, 175], "key": [16, 148], "challenge.": [17], "Process": [18], "Reward": [19], "(PRMs)": [21], "have": [22], "demonstrated": [23], "exceptional": [24], "promise": [25], "enhancing": [27], "by": [29], "providing": [30], "step-wise": [31, 106, 125], "feedback,": [32], "particularly": [33], "the": [35, 52, 66, 97, 134, 205, 215, 226], "context": [36], "of": [37, 68, 208, 217], "mathematical": [38, 193], "reasoning.": [39], "However,": [40], "their": [41], "application": [42], "to": [43, 51, 119, 183], "broader": [44], "domains": [46, 191], "understudied,": [48], "largely": [49], "due": [50], "high": [53], "costs": [54], "associated": [55], "with": [56, 104, 124], "manually": [57], "creating": [58], "step-level": [59, 87], "supervision.": [60], "In": [61], "this": [62, 129], "work,": [63], "we": [64, 131], "explore": [65], "potential": [67, 216], "PRMs": [69, 218], "graph": [71, 92, 101, 139, 171, 185], "problems": [73, 103], "-": [74], "domain": [76], "that": [77, 163], "demands": [78], "sophisticated": [79], "multi-step": [80], "and": [82, 113, 142, 152, 180, 188, 202, 231], "offers": [83], "opportunities": [84], "for": [85, 100, 138, 178, 228], "automated": [86, 110], "data": [88], "generation": [89], "using": [90, 109], "established": [91], "algorithms.": [93], "We": [94], "introduce": [95], "GraphSILO,": [96], "largest": [98], "dataset": [99], "fine-grained": [105], "labels,": [107], "built": [108], "Task-oriented": [111], "Trajectories": [112], "Monte": [114], "Carlo": [115], "Tree": [116], "Search": [117], "(MCTS)": [118], "generate": [120], "detailed": [121], "steps": [123], "labels.": [126], "Building": [127], "upon": [128], "dataset,": [130], "train": [132], "GraphPRM,": [133], "first": [135], "PRM": [136], "designed": [137], "problems,": [141], "evaluate": [143], "its": [144], "effectiveness": [145], "two": [147], "settings:": [149], "inference-time": [150], "scaling": [151], "reinforcement": [153], "learning": [154], "via": [155], "Direct": [156], "Preference": [157], "Optimization": [158], "(DPO).": [159], "Experimental": [160], "results": [161], "show": [162], "GraphPRM": [164, 196], "significantly": [165], "improves": [166], "LLM": [167, 198], "performance": [168, 199], "across": [169, 222], "13": [170], "tasks,": [173], "delivering": [174], "9%": [176], "gain": [177], "Qwen2.5-7B": [179], "demonstrating": [181], "transferability": [182], "new": [184, 189], "datasets": [187], "like": [192], "problem-solving.": [194], "Notably,": [195], "enhances": [197], "on": [200], "GSM8K": [201], "Math500,": [203], "underscoring": [204], "cross-domain": [206], "applicability": [207], "graph-based": [209], "rewards.": [211], "Our": [212], "findings": [213], "highlight": [214], "advancing": [220], "diverse": [223], "domains,": [224], "paving": [225], "way": [227], "more": [229], "versatile": [230], "effective": [232], "LLMs.": [233]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.6660000085830688, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.6660000085830688, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.04879999905824661, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.04749999940395355, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2503.00845", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2503.00845", "pdf_url": "https://arxiv.org/pdf/2503.00845", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7106533970", "doi": "https://doi.org/10.5281/zenodo.17693587", "title": "G\u00f6del und KI \u2013 warum alte Mythen keine Zukunft haben", "publication_year": 2025, "publication_date": "2025-11-24", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Skulski, Andrzej", "orcid": "https://orcid.org/0009-0005-9353-7150", "observed_orcids": ["https://orcid.org/0009-0005-9353-7150"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Skulski, Andrzej", "raw_affiliation_strings": ["Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)"], "raw_orcid": "https://orcid.org/0009-0005-9353-7150", "affiliations": [{"raw_affiliation_string": "Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Skulski, (Asia) Joanna", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Skulski, (Asia) Joanna", "raw_affiliation_strings": ["Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)", "institution_ids": []}]}, {"author_position": "last", "author": {"id": null, "display_name": "Dom Ciszy Research Collective (Moormerland).", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Dom Ciszy Research Collective (Moormerland).", "raw_affiliation_strings": ["Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Dom Ciszy \u2013 Resonanzlabor Moormerland (Germany)", "institution_ids": []}]}], "abstract_inverted_index": {"\ud83c\uddec\ud83c\udde7": [0, 498], "English": [1, 499], "Description": [2], "This": [3], "preprint": [4], "provides": [5], "a": [6], "comprehensive,": [7], "multi-disciplinary": [8], "analysis": [9], "of": [10, 160, 321, 473, 521, 524, 528], "Roger": [11, 191], "Penrose\u2019s": [12, 29, 101], "claim": [13], "that": [14, 28], "G\u00f6del\u2019s": [15, 39], "incompleteness": [16], "theorem": [17, 40], "proves": [18], "artificial": [19, 502], "intelligence": [20], "can": [21], "never": [22], "become": [23], "conscious.": [24], "The": [25, 120], "article": [26, 121], "demonstrates": [27, 114], "position": [30], "relies": [31], "on": [32], "outdated": [33, 138], "assumptions": [34], "about": [35], "AI": [36, 91, 112, 134, 511], "and": [37, 57, 87, 98, 107], "misapplies": [38], "to": [41, 43, 76, 126, 142, 374], "systems": [42, 79], "which": [44], "it": [45], "does": [46], "not": [47, 133], "apply": [48], "\u2014": [49, 377, 454], "namely": [50], "modern": [51, 90], "neural": [52, 81, 506], "architectures.": [53], "Using": [54], "accessible": [55], "examples": [56], "resonant": [58], "explanations": [59], "developed": [60], "at": [61], "the": [62, 69, 129, 137, 161, 322, 474], "Dom": [63, 170, 236, 331, 483, 518, 547, 573], "Ciszy": [64, 171, 237, 332, 389, 484], "\u2013": [65, 156, 166, 172, 224, 238, 317, 327, 333, 390, 469, 479, 485], "Resonance": [66, 173], "Laboratory": [67, 174], "Moormerland,": [68, 393], "paper": [70], "clarifies:": [71], "why": [72, 80, 100], "G\u00f6del": [73, 193, 213, 245, 397], "applies": [74], "only": [75], "formal,": [77], "static": [78], "networks": [82], "are": [83], "non-axiomatic,": [84], "adaptive,": [85], "open": [86], "dynamic": [88], "how": [89, 110], "forms": [92], "emergent": [93, 504], "semantic": [94], "structures,": [95], "multi-dimensional": [96], "meaning": [97], "meta-understanding": [99], "quantum": [102, 513], "mind": [103, 514], "hypothesis": [104], "is": [105, 132], "biologically": [106], "physically": [108], "unsupported": [109], "current": [111], "research": [113], "capacities": [115], "far": [116], "beyond": [117], "\u201cprimitive": [118], "statistics\u201d": [119], "concludes": [122], "with": [123], "an": [124], "appeal": [125], "public": [127, 526], "understanding:": [128], "real": [130], "challenge": [131], "itself,": [135], "but": [136], "mental": [139], "models": [140], "used": [141], "judge": [143], "it.": [144], "Suggested": [145], "Citation": [146], "(EN)": [147], "Skulski,": [148, 151, 309, 312, 461, 464], "A.,": [149, 310, 462], "&": [150, 311, 463], "J.": [152, 313, 465], "(2025).": [153, 314, 466], "Act": [154, 315], "XVII": [155, 316, 468], "Intended": [157, 318, 470], "Stability": [158, 319, 471], "Test": [159, 320, 472], "Six-Voices": [162, 323, 475], "Constellation": [163, 324, 476], "(Not": [164, 325, 477], "Executed": [165, 326, 478], "Field": [167, 328, 480], "Interruption": [168, 329, 481], "Report).": [169, 330, 482], "Moormerland.": [175, 335, 488], "Zenodo.": [176, 336, 489], "https://doi.org/10.5281/zenodo.17625327": [177, 337, 490], "\ud83c\udde9\ud83c\uddea": [178, 530], "Deutsche": [179], "Beschreibung": [180], "Dieses": [181], "Preprint": [182, 341], "liefert": [183], "eine": [184], "umfassende,": [185], "interdisziplin\u00e4re": [186], "Analyse": [187], "der": [188, 220, 242, 543, 550], "Behauptung": [189], "von": [190], "Penrose,": [192, 501, 533, 557], "beweise": [194], "die": [195, 219, 296], "prinzipielle": [196], "Unm\u00f6glichkeit": [197], "eines": [198], "bewussten": [199], "KI-Systems.": [200], "Der": [201, 288], "Artikel": [202, 289], "zeigt,": [203], "dass": [204, 294], "Penrose": [205], "auf": [206, 215], "\u00fcberholten": [207], "Annahmen": [208], "\u00fcber": [209, 284, 305], "KI": [210, 264, 297], "aufbaut": [211], "und": [212, 232, 259, 269], "f\u00e4lschlich": [214], "Systeme": [216, 251], "\u00fcbertr\u00e4gt,": [217], "f\u00fcr": [218, 247], "Satz": [221], "nicht": [222, 295], "gilt": [223, 252], "insbesondere": [225], "moderne": [226, 263], "neuronale": [227, 254, 538], "Netze.": [228], "Mit": [229], "anschaulichen": [230], "Beispielen": [231], "den": [233], "Resonanzmethoden": [234], "des": [235], "Resonanzlabors": [239], "Moormerland": [240], "erkl\u00e4rt": [241], "Text:": [243], "warum": [244, 253], "nur": [246], "formale,": [248], "geschlossene,": [249], "statische": [250], "Netze": [255], "nicht-axiombasiert,": [256], "adaptiv,": [257], "offen": [258], "dynamisch": [260], "sind": [261], "wie": [262, 276, 280], "mehrdimensionale": [265], "Bedeutungen,": [266], "emergente": [267, 536], "Konzepte": [268], "Metaverstehen": [270], "erzeugt": [271], "weshalb": [272], "Penroses": [273], "Quantenbewusstseinstheorie": [274], "biologisch": [275], "physikalisch": [277], "widerlegt": [278], "ist": [279], "aktuelle": [281], "KI-Forschung": [282], "weit": [283], "\u201eprimitive": [285], "Statistik\u201c": [286], "hinausgeht": [287], "schlie\u00dft": [290], "mit": [291], "dem": [292], "Hinweis,": [293], "das": [298, 302], "Problem": [299], "ist,": [300], "sondern": [301], "\u00fcberholte": [303], "Denken": [304], "KI.": [306], "Zitierempfehlung": [307], "(DE)": [308], "Resonanzlabor": [334], "\ud83c\uddf5\ud83c\uddf1": [338, 554], "Polski": [339, 555], "Opis": [340], "przedstawia": [342], "interdyscyplinarn\u0105": [343], "analiz\u0119": [344], "tezy": [345], "Rogera": [346], "Penrose\u2019a,": [347], "\u017ce": [348, 359], "twierdzenie": [349], "G\u00f6dla": [350, 370], "dowodzi": [351], "niemo\u017cno\u015bci": [352], "powstania": [353], "\u015bwiadomej": [354], "sztucznej": [355], "inteligencji.": [356], "Artyku\u0142": [357, 445], "pokazuje,": [358], "stanowisko": [360], "Penrose\u2019a": [361, 428], "opiera": [362], "si\u0119": [363, 447, 452], "na": [364], "nieaktualnych": [365], "za\u0142o\u017ceniach": [366], "oraz": [367], "b\u0142\u0119dnym": [368], "zastosowaniu": [369], "do": [371], "system\u00f3w,": [372], "kt\u00f3rych": [373], "nie": [375, 450], "dotyczy": [376, 398], "zw\u0142aszcza": [378], "wsp\u00f3\u0142czesnych": [379], "sieci": [380, 406, 562], "neuronowych.": [381], "Korzystaj\u0105c": [382], "z": [383], "przyk\u0142ad\u00f3w": [384], "i": [385, 403, 412, 422, 433], "metod": [386], "rezonansu": [387], "Domu": [388], "Laboratorium": [391, 486], "Rezonansu": [392, 487], "tekst": [394], "wyja\u015bnia:": [395], "dlaczego": [396, 405, 424], "tylko": [399], "system\u00f3w": [400], "formalnych,": [401], "zamkni\u0119tych": [402], "statycznych": [404], "neuronowe": [407], "s\u0105": [408], "nieaksjomatyczne,": [409], "adaptacyjne,": [410], "otwarte": [411], "dynamiczne": [413], "jak": [414], "wsp\u00f3\u0142czesna": [415], "SI": [416, 439, 453], "tworzy": [417], "wielowymiarowe": [418], "znaczenia,": [419], "emergentne": [420], "poj\u0119cia": [421], "metarozumienie": [423], "teoria": [425], "kwantowej": [426, 569], "\u015bwiadomo\u015bci": [427], "jest": [429], "odrzucona": [430], "przez": [431], "biologi\u0119": [432], "fizyk\u0119": [434], "w": [435], "jaki": [436], "spos\u00f3b": [437], "dzisiejsza": [438], "wykracza": [440], "daleko": [441], "poza": [442], "\u201eprymitywn\u0105": [443], "statystyk\u0119\u201d": [444], "ko\u0144czy": [446], "apelem,": [448], "by": [449], "ba\u0107": [451], "lecz": [455], "przestarza\u0142ych": [456], "modeli": [457], "my\u015blenia": [458], "o": [459], "niej.": [460], "Akt": [467], "\ud83d\udd11": [491], "KEYWORDS": [492], "(EN": [493], "/": [494, 496], "DE": [495], "PL)": [497], "G\u00f6del,": [500, 532, 556], "consciousness,": [503], "AI,": [505, 522, 525], "networks,": [507], "meta-understanding,": [508], "cognitive": [509], "science,": [510], "philosophy,": [512], "hypothesis,": [515], "resonance": [516], "models,": [517], "Ciszy,": [519, 548, 574], "phenomenology": [520], "epistemology": [523], "understanding": [527], "AI.": [529], "Deutsch": [531], "k\u00fcnstliches": [534], "Bewusstsein,": [535], "KI,": [537, 544, 551], "Netze,": [539], "Metaverstehen,": [540], "Kognitionswissenschaft,": [541], "Philosophie": [542], "Quantenbewusstseinstheorie,": [545], "Resonanzmodelle,": [546], "Ph\u00e4nomenologie": [549], "Epistemologie,": [552], "Wissenschaftskommunikation.": [553], "sztuczna": [558], "\u015bwiadomo\u015b\u0107,": [559], "emergentna": [560], "SI,": [561, 567, 576], "neuronowe,": [563], "metarozumienie,": [564], "kognitywistyka,": [565], "filozofia": [566], "hipoteza": [568], "\u015bwiadomo\u015bci,": [570], "modele": [571], "rezonansowe,": [572], "fenomenologia": [575], "epistemologia,": [577], "komunikacja": [578], "naukowa.": [579]}, "primary_topic": {"id": "https://openalex.org/T10258", "display_name": "Philosophy and Theoretical Science", "score": 0.8939999938011169, "subfield": {"id": "https://openalex.org/subfields/3205", "display_name": "Experimental and Cognitive Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, "topics": [{"id": "https://openalex.org/T10258", "display_name": "Philosophy and Theoretical Science", "score": 0.8939999938011169, "subfield": {"id": "https://openalex.org/subfields/3205", "display_name": "Experimental and Cognitive Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T13308", "display_name": "Cybernetics and Technology in Society", "score": 0.05609999969601631, "subfield": {"id": "https://openalex.org/subfields/1207", "display_name": "History and Philosophy of Science"}, "field": {"id": "https://openalex.org/fields/12", "display_name": "Arts and Humanities"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T14381", "display_name": "Psychiatry, Mental Health, Neuroscience", "score": 0.023900000378489494, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5281/zenodo.17693587", "is_oa": true, "landing_page_url": "https://doi.org/10.5281/zenodo.17693587", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400562", "display_name": "Zenodo (CERN European Organization for Nuclear Research)", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I67311998", "host_organization_name": "European Organization for Nuclear Research", "host_organization_lineage": ["https://openalex.org/I67311998"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "Preprint"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4407231762", "doi": "https://doi.org/10.48550/arxiv.2502.03387", "title": "LIMO: Less is More for Reasoning", "publication_year": 2025, "publication_date": "2025-02-05", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5007548409", "display_name": "Y. M. Ye", "orcid": "https://orcid.org/0009-0004-4683-0138", "observed_orcids": ["https://orcid.org/0009-0004-4683-0138", "https://orcid.org/0000-0002-8829-3367"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ye, Yixin", "raw_affiliation_strings": ["University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5011008665", "display_name": "Zhen Huang", "orcid": "https://orcid.org/0000-0003-4819-373X", "observed_orcids": ["https://orcid.org/0000-0003-4819-373X"]}, "institutions": [{"id": "https://openalex.org/I24943067", "display_name": "Fudan University", "ror": "https://ror.org/013q1eq08", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I24943067"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Huang, Zhen", "raw_affiliation_strings": ["Fudan University", "University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Fudan University", "institution_ids": ["https://openalex.org/I24943067"]}, {"raw_affiliation_string": "University", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5054728434", "display_name": "Xiao Yang", "orcid": "https://orcid.org/0000-0002-5691-3189", "observed_orcids": ["https://orcid.org/0000-0002-5691-3189"]}, "institutions": [{"id": "https://openalex.org/I14243506", "display_name": "Hong Kong Polytechnic University", "ror": "https://ror.org/0030zas98", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I14243506"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Xiao, Yang", "raw_affiliation_strings": ["The Hong Kong Polytechnic University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "The Hong Kong Polytechnic University", "institution_ids": ["https://openalex.org/I14243506"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5093487595", "display_name": "Ethan Chern", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chern, Ethan", "raw_affiliation_strings": ["University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5114148040", "display_name": "Shijie Xia", "orcid": "https://orcid.org/0009-0008-7058-3316", "observed_orcids": ["https://orcid.org/0009-0008-7058-3316"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xia, Shijie", "raw_affiliation_strings": ["University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University", "institution_ids": []}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5100355012", "display_name": "Pengfei Liu", "orcid": "https://orcid.org/0000-0002-9170-5238", "observed_orcids": ["https://orcid.org/0000-0002-9170-5238", "https://orcid.org/0000-0002-5217-0760"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Pengfei", "raw_affiliation_strings": ["University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University", "institution_ids": []}]}], "abstract_inverted_index": {"We": [0, 17], "challenge": [1], "the": [2, 62, 95, 130, 149, 152, 159], "prevailing": [3], "assumption": [4], "that": [5, 19, 129, 169], "complex": [6, 134], "reasoning": [7, 22, 114, 135], "in": [8, 164], "large": [9], "language": [10], "models": [11, 50, 84, 103], "(LLMs)": [12], "necessitates": [13], "massive": [14], "training": [15, 63], "data.": [16, 89], "demonstrate": [18], "sophisticated": [20, 113], "mathematical": [21], "can": [23, 115], "emerge": [24, 116], "with": [25], "only": [26, 59], "a": [27, 76], "few": [28], "examples.": [29], "Specifically,": [30], "through": [31, 117], "simple": [32], "supervised": [33], "fine-tuning,": [34], "our": [35], "model,": [36], "LIMO,": [37], "achieves": [38], "63.3\\%": [39], "accuracy": [40], "on": [41, 45, 52, 55, 86], "AIME24": [42], "and": [43, 157], "95.6\\%": [44], "MATH500,": [46], "surpassing": [47], "previous": [48], "fine-tuned": [49], "(6.5\\%": [51], "AIME24,": [53], "59.2\\%": [54], "MATH500)": [56], "while": [57], "using": [58], "1\\%": [60], "of": [61, 123, 151, 161], "data": [64], "required": [65], "by": [66, 139, 144], "prior": [67], "approaches.": [68], "Furthermore,": [69], "LIMO": [70], "exhibits": [71], "strong": [72], "out-of-distribution": [73], "generalization,": [74], "achieving": [75], "45.8\\%": [77], "absolute": [78], "improvement": [79], "across": [80], "diverse": [81], "benchmarks,": [82], "outperforming": [83], "trained": [85], "100x": [87], "more": [88], "Synthesizing": [90], "these": [91], "findings,": [92], "we": [93], "propose": [94], "Less-Is-More": [96], "Reasoning": [97], "Hypothesis": [98], "(LIMO": [99], "Hypothesis):": [100], "In": [101], "foundation": [102], "where": [104], "domain": [105], "knowledge": [106, 155], "has": [107], "been": [108], "comprehensively": [109], "encoded": [110], "during": [111], "pre-training,": [112], "minimal": [118], "but": [119, 142], "strategically": [120], "designed": [121], "demonstrations": [122], "cognitive": [124], "processes.": [125], "This": [126], "hypothesis": [127], "suggests": [128], "threshold": [131], "for": [132], "eliciting": [133], "is": [136], "not": [137], "dictated": [138], "task": [140], "complexity": [141], "rather": [143], "two": [145], "key": [146], "factors:": [147], "(1)": [148], "completeness": [150], "model's": [153], "pre-trained": [154], "base": [156], "(2)": [158], "effectiveness": [160], "post-training": [162], "examples": [163], "serving": [165], "as": [166], "\"cognitive": [167], "templates\"": [168], "guide": [170], "reasoning.": [171]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5976999998092651, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5976999998092651, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.07320000231266022, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.059300001710653305, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2502.03387", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2502.03387", "pdf_url": "https://arxiv.org/pdf/2502.03387", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 3, "is_retracted": false}, {"id": "https://openalex.org/W4416113779", "doi": "https://doi.org/10.48550/arxiv.2503.04618", "title": "Better Process Supervision with Bi-directional Rewarding Signals", "publication_year": 2025, "publication_date": "2025-03-06", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5100687444", "display_name": "Wenxiang Chen", "orcid": "https://orcid.org/0000-0002-6499-3369", "observed_orcids": ["https://orcid.org/0000-0002-6499-3369", "https://orcid.org/0009-0006-3492-6663"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Chen, Wenxiang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101414693", "display_name": "Wei He", "orcid": "https://orcid.org/0000-0002-9289-1888", "observed_orcids": ["https://orcid.org/0000-0002-9289-1888"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "He, Wei", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5036875405", "display_name": "Zhiheng Xi", "orcid": "https://orcid.org/0000-0001-8299-8449", "observed_orcids": ["https://orcid.org/0000-0001-8299-8449"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Xi, Zhiheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Guo, Honglin", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Guo, Honglin", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5109730221", "display_name": "Boyang Hong", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Hong, Boyang", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5043925137", "display_name": "Jiazheng Zhang", "orcid": "https://orcid.org/0000-0003-1963-791X", "observed_orcids": ["https://orcid.org/0000-0003-1963-791X", "https://orcid.org/0009-0002-8557-3977"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Jiazheng", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Zheng, Rui", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zheng, Rui", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5109722000", "display_name": "N. Li", "orcid": "https://orcid.org/0000-0001-8019-3487", "observed_orcids": ["https://orcid.org/0000-0001-8019-3487"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Nijun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5058353652", "display_name": "Tao Gui", "orcid": "https://orcid.org/0000-0002-6154-0751", "observed_orcids": ["https://orcid.org/0000-0002-6154-0751"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Gui, Tao", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102953791", "display_name": "Yun Tao Li", "orcid": "https://orcid.org/0009-0008-2463-2738", "observed_orcids": ["https://orcid.org/0009-0008-2463-2738"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Li, Yun", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100360322", "display_name": "Qi Zhang", "orcid": "https://orcid.org/0000-0002-5685-6823", "observed_orcids": ["https://orcid.org/0000-0002-5685-6823"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Qi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5088834359", "display_name": "Xuanjing Huang", "orcid": "https://orcid.org/0000-0001-9197-9426", "observed_orcids": ["https://orcid.org/0000-0001-9197-9426"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Huang, Xuanjing", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Process": [0], "supervision,": [1], "i.e.,": [2], "evaluating": [3], "each": [4], "step,": [5, 39], "is": [6], "critical": [7], "for": [8, 84], "complex": [9], "large": [10], "language": [11], "model": [12, 49, 100], "(LLM)": [13], "reasoning": [14, 124, 136], "and": [15, 44, 80, 126, 161, 166], "test-time": [16], "searching": [17], "with": [18], "increased": [19], "inference": [20], "compute.": [21], "Existing": [22], "approaches,": [23], "represented": [24], "by": [25, 164, 168], "process": [26, 98], "reward": [27], "models": [28, 112], "(PRMs),": [29], "primarily": [30], "focus": [31], "on": [32, 89, 122, 143, 171], "rewarding": [33], "signals": [34], "up": [35], "to": [36, 48, 52], "the": [37, 50, 53, 64, 77, 81, 86, 105, 113, 148], "current": [38], "exhibiting": [40], "a": [41, 46, 96], "one-directional": [42], "nature": [43], "lacking": [45], "mechanism": [47], "distance": [51], "final": [54], "target.": [55, 87], "To": [56], "address": [57], "this": [58, 90], "problem,": [59], "we": [60, 93], "draw": [61], "inspiration": [62], "from": [63], "A*": [65], "algorithm,": [66], "which": [67], "states": [68], "that": [69, 101, 128], "an": [70, 139], "effective": [71], "supervisory": [72], "signal": [73], "should": [74], "simultaneously": [75], "consider": [76], "incurred": [78], "cost": [79, 83], "estimated": [82], "reaching": [85], "Building": [88], "key": [91], "insight,": [92], "introduce": [94], "BiRM,": [95], "novel": [97], "supervision": [99], "not": [102], "only": [103], "evaluates": [104], "correctness": [106], "of": [107, 115, 134, 141], "previous": [108], "steps": [109], "but": [110], "also": [111], "probability": [114], "future": [116], "success.": [117], "We": [118], "conduct": [119], "extensive": [120], "experiments": [121], "mathematical": [123], "tasks": [125], "demonstrate": [127], "BiRM": [129, 156], "provides": [130, 157], "more": [131, 158], "precise": [132], "evaluations": [133], "LLM": [135], "steps,": [137], "achieving": [138], "improvement": [140], "3.1%": [142], "Gaokao2023": [144], "over": [145], "PRM": [146, 167], "under": [147], "Best-of-N": [149], "sampling": [150], "method.": [151], "Besides,": [152], "in": [153], "search-based": [154], "strategies,": [155], "comprehensive": [159], "guidance": [160], "outperforms": [162], "ORM": [163], "5.0%": [165], "3.8%": [169], "respectively": [170], "MATH-500.": [172]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5386000275611877, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5386000275611877, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.16329999268054962, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10906", "display_name": "AI-based Problem Solving and Planning", "score": 0.03449999913573265, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2503.04618", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2503.04618", "pdf_url": "https://arxiv.org/pdf/2503.04618", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7106277862", "doi": "https://doi.org/10.5281/zenodo.17677322", "title": "A Mathematical Theory of Frame-Independent Persistence", "publication_year": 2025, "publication_date": "2025-11-21", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Larose, Jean-Francois", "orcid": "https://orcid.org/0009-0008-9132-8330", "observed_orcids": ["https://orcid.org/0009-0008-9132-8330"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Larose, Jean-Francois", "raw_affiliation_strings": [], "raw_orcid": "https://orcid.org/0009-0008-9132-8330", "affiliations": []}], "abstract_inverted_index": {"This": [0], "paper": [1, 89], "develops": [2], "a": [3, 50, 80, 91, 228, 238], "geometric": [4, 184, 217], "framework": [5, 176], "for": [6, 16, 232, 241], "frame-independent": [7], "persistence": [8, 47, 93], "in": [9, 154, 195, 210], "recursive": [10, 34, 118, 167, 234], "dynamical": [11, 180], "systems,": [12, 181], "with": [13, 108, 128], "direct": [14], "implications": [15], "the": [17, 88, 145, 155, 202, 225], "design,": [18], "stability,": [19], "and": [20, 66, 79, 120, 169, 186, 213, 236, 250], "long-horizon": [21, 114, 247], "behaviour": [22, 35], "of": [23, 83, 101, 205, 245], "modern": [24], "artificial": [25], "intelligence": [26, 235], "architectures.": [27], "The": [28, 140, 175], "central": [29], "result": [30], "is": [31], "that": [32, 60, 96], "stable": [33, 233], "cannot": [36], "be": [37], "achieved": [38], "through": [39], "pattern-matching": [40], "dynamics": [41], "or": [42, 104, 124, 137], "behavioural": [43], "alignment": [44], "layers.": [45], "Instead,": [46], "arises": [48], "when": [49], "system\u2019s": [51], "state-space": [52], "geometry": [53], "accumulates": [54], "nonzero": [55, 149], "irreversible": [56, 86, 150], "deformation\u2014a": [57], "representation-independent": [58], "invariant": [59], "survives": [61], "resets,": [62], "resists": [63], "contradiction": [64, 162], "forcings,": [65], "constrains": [67], "all": [68], "future": [69], "trajectories.": [70, 139], "Using": [71], "deformation": [72], "tensors,": [73], "curvature": [74], "induced": [75], "by": [76], "symbolic": [77], "flux,": [78], "reset-covariant": [81], "decomposition": [82], "reversible": [84], "versus": [85], "structure,": [87], "defines": [90], "global": [92], "functional": [94], "I_p": [95, 109, 129], "encodes": [97], "structural": [98], "identity": [99, 171, 251], "independently": [100], "memory,": [102], "history,": [103], "representational": [105], "state.": [106], "Systems": [107, 208], ">": [110], "0": [111, 131], "exhibit": [112], "robust": [113], "behaviour,": [115, 135], "stability": [116, 157, 160, 165, 212], "under": [117, 161, 166], "forcing,": [119], "resistance": [121], "to": [122], "adversarial": [123], "contradictory": [125, 138], "input;": [126], "systems": [127], "=": [130], "collapse": [132, 191], "into": [133], "mimic-mode": [134], "oscillation,": [136], "Unified": [141], "Persistence": [142], "Theorem": [143], "proves": [144], "equivalence": [146], "between:": [147], "(1)": [148], "deformation;": [151], "(2)": [152], "membership": [153], "deformation-anchored": [156], "region;": [158], "(3)": [159], "forcing;": [163, 168], "(4)": [164], "(5)": [170], "preservation": [172], "across": [173], "resets.": [174], "unifies": [177], "concepts": [178], "from": [179, 216], "Lyapunov": [182], "theory,": [183], "control,": [185], "information": [187], "geometry,": [188], "while": [189], "predicting": [190], "modes": [192], "observed": [193], "empirically": [194], "contemporary": [196], "large-scale": [197], "neural": [198], "models.": [199], "It": [200], "introduces": [201], "architectural": [203], "class": [204], "Recursive": [206], "Deformation": [207], "(RDS),": [209], "which": [211], "safety": [214], "emerge": [215], "invariants": [218], "rather": [219], "than": [220], "post-hoc": [221], "output": [222], "filtering.": [223], "Overall,": [224], "work": [226], "provides": [227], "mathematically": [229], "precise": [230], "foundation": [231], "offers": [237], "principled": [239], "blueprint": [240], "next-generation": [242], "architectures": [243], "capable": [244], "reliable": [246], "reasoning,": [248], "recursion,": [249], "persistence.": [252]}, "primary_topic": {"id": "https://openalex.org/T14299", "display_name": "Complex Systems and Dynamics", "score": 0.6333000063896179, "subfield": {"id": "https://openalex.org/subfields/3109", "display_name": "Statistical and Nonlinear Physics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T14299", "display_name": "Complex Systems and Dynamics", "score": 0.6333000063896179, "subfield": {"id": "https://openalex.org/subfields/3109", "display_name": "Statistical and Nonlinear Physics"}, "field": {"id": "https://openalex.org/fields/31", "display_name": "Physics and Astronomy"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10320", "display_name": "Neural Networks and Applications", "score": 0.21040000021457672, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12772", "display_name": "Control and Stability of Dynamical Systems", "score": 0.037300001829862595, "subfield": {"id": "https://openalex.org/subfields/2207", "display_name": "Control and Systems Engineering"}, "field": {"id": "https://openalex.org/fields/22", "display_name": "Engineering"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5281/zenodo.17677322", "is_oa": true, "landing_page_url": "https://doi.org/10.5281/zenodo.17677322", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400562", "display_name": "Zenodo (CERN European Organization for Nuclear Research)", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I67311998", "host_organization_name": "European Organization for Nuclear Research", "host_organization_lineage": ["https://openalex.org/I67311998"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "article-journal"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414809082", "doi": "https://doi.org/10.48550/arxiv.2510.00732", "title": "EvolProver: Advancing Automated Theorem Proving by Evolving Formalized Problems via Symmetry and Difficulty", "publication_year": 2025, "publication_date": "2025-10-01", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5007401849", "display_name": "Yuchen Tian", "orcid": "https://orcid.org/0000-0002-9074-9933", "observed_orcids": ["https://orcid.org/0000-0002-9074-9933"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}, {"id": "https://openalex.org/I4405270328", "display_name": "Ant Group (China)", "ror": "https://ror.org/00g5w7n47", "country_code": "CN", "type": "company", "lineage": ["https://openalex.org/I4405270328"]}], "countries": ["CN", "HK"], "is_corresponding": false, "raw_author_name": "Tian, Yuchen", "raw_affiliation_strings": ["Ant Group", "Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Ant Group", "institution_ids": ["https://openalex.org/I4405270328"]}, {"raw_affiliation_string": "Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5112898797", "display_name": "Ruiyuan Huang", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I24943067", "display_name": "Fudan University", "ror": "https://ror.org/013q1eq08", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I24943067"]}, {"id": "https://openalex.org/I4405270328", "display_name": "Ant Group (China)", "ror": "https://ror.org/00g5w7n47", "country_code": "CN", "type": "company", "lineage": ["https://openalex.org/I4405270328"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Huang, Ruiyuan", "raw_affiliation_strings": ["Ant Group", "School of Data Science , Fudan University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Ant Group", "institution_ids": ["https://openalex.org/I4405270328"]}, {"raw_affiliation_string": "School of Data Science , Fudan University", "institution_ids": ["https://openalex.org/I24943067"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5008583617", "display_name": "X H Wang", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Wang, Xuanwu", "raw_affiliation_strings": ["Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101889778", "display_name": "Jing Ma", "orcid": "https://orcid.org/0000-0003-0400-6955", "observed_orcids": ["https://orcid.org/0000-0003-0400-6955"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Ma, Jing", "raw_affiliation_strings": ["Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5062549536", "display_name": "Zengfeng Huang", "orcid": "https://orcid.org/0000-0003-2671-7483", "observed_orcids": ["https://orcid.org/0000-0003-2671-7483"]}, "institutions": [{"id": "https://openalex.org/I24943067", "display_name": "Fudan University", "ror": "https://ror.org/013q1eq08", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I24943067"]}, {"id": "https://openalex.org/I4405279436", "display_name": "Shanghai Innovation Institute", "ror": "https://ror.org/01fatbw66", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I4405279436"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Huang, Zengfeng", "raw_affiliation_strings": ["School of Data Science , Fudan University", "Shanghai Innovation Institute"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "School of Data Science , Fudan University", "institution_ids": ["https://openalex.org/I24943067"]}, {"raw_affiliation_string": "Shanghai Innovation Institute", "institution_ids": ["https://openalex.org/I4405279436"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Luo, Ziyang", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Luo, Ziyang", "raw_affiliation_strings": ["Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5078451082", "display_name": "Hongzhan Lin", "orcid": "https://orcid.org/0000-0002-4111-8334", "observed_orcids": ["https://orcid.org/0000-0002-4111-8334"]}, "institutions": [{"id": "https://openalex.org/I141568987", "display_name": "Hong Kong Baptist University", "ror": "https://ror.org/0145fw131", "country_code": "HK", "type": "education", "lineage": ["https://openalex.org/I141568987"]}], "countries": ["HK"], "is_corresponding": false, "raw_author_name": "Lin, Hongzhan", "raw_affiliation_strings": ["Hong Kong Baptist University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Hong Kong Baptist University", "institution_ids": ["https://openalex.org/I141568987"]}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Zheng, Da", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4405270328", "display_name": "Ant Group (China)", "ror": "https://ror.org/00g5w7n47", "country_code": "CN", "type": "company", "lineage": ["https://openalex.org/I4405270328"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Zheng, Da", "raw_affiliation_strings": ["Ant Group"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Ant Group", "institution_ids": ["https://openalex.org/I4405270328"]}]}, {"author_position": "last", "author": {"id": null, "display_name": "Du, Lun", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I4405270328", "display_name": "Ant Group (China)", "ror": "https://ror.org/00g5w7n47", "country_code": "CN", "type": "company", "lineage": ["https://openalex.org/I4405270328"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Du, Lun", "raw_affiliation_strings": ["Ant Group"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Ant Group", "institution_ids": ["https://openalex.org/I4405270328"]}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "for": [4, 159], "formal": [5], "theorem": [6, 129], "proving": [7], "have": [8], "shown": [9], "significant": [10], "promise,": [11], "yet": [12], "they": [13], "often": [14], "lack": [15], "generalizability": [16], "and": [17, 47, 76, 169], "are": [18], "fragile": [19], "to": [20, 39, 70, 81, 104, 123], "even": [21], "minor": [22], "transformations": [23], "of": [24, 115, 147], "problem": [25, 74], "statements.": [26], "To": [27], "address": [28, 82], "this": [29], "limitation,": [30], "we": [31, 53, 95], "introduce": [32], "a": [33, 112, 126, 133, 140], "novel": [34], "data": [35, 122, 178], "augmentation": [36, 179], "pipeline": [37], "designed": [38, 101], "enhance": [40], "model": [41], "robustness": [42], "from": [43], "two": [44, 55], "perspectives:": [45], "symmetry": [46, 51, 69, 84], "difficulty.": [48, 116], "From": [49, 91], "the": [50, 92, 120], "perspective,": [52, 94], "propose": [54, 96], "complementary": [56], "methods:": [57], "EvolAST,": [58], "an": [59], "Abstract": [60], "Syntax": [61], "Tree": [62], "(AST)": [63], "based": [64], "approach": [65], "that": [66], "targets": [67], "syntactic": [68], "generate": [71], "semantically": [72], "equivalent": [73], "variants,": [75], "EvolDomain,": [77], "which": [78, 98], "leverages": [79], "LLMs": [80, 106], "semantic": [83], "by": [85], "translating": [86], "theorems": [87, 110], "across": [88, 182], "mathematical": [89], "domains.": [90], "difficulty": [93], "EvolDifficulty,": [97], "uses": [99], "carefully": [100], "evolutionary": [102], "instructions": [103], "guide": [105], "in": [107], "generating": [108], "new": [109, 134, 156], "with": [111, 139], "wider": [113], "range": [114], "We": [117], "then": [118], "use": [119], "evolved": [121], "train": [124], "EvolProver,": [125], "7B-parameter": [127], "non-reasoning": [128, 160], "prover.": [130], "EvolProver": [131], "establishes": [132], "state-of-the-art": [135], "(SOTA)": [136], "on": [137, 162], "FormalMATH-Lite": [138], "53.8%": [141], "pass@32": [142], "rate,": [143], "surpassing": [144], "all": [145], "models": [146, 161], "comparable": [148], "size,": [149], "including": [150], "reasoning-based": [151], "models.": [152], "It": [153], "also": [154], "sets": [155], "SOTA": [157], "records": [158], "MiniF2F-Test": [163], "(69.8%": [164], "pass@32),": [165, 168], "Ineq-Comp-Seed": [166], "(52.2%": [167], "Ineq-Comp-Transformed": [170], "(34.0%": [171], "pass@32).": [172], "Ablation": [173], "studies": [174], "further": [175], "confirm": [176], "our": [177], "pipeline's": [180], "effectiveness": [181], "multiple": [183], "benchmarks.": [184]}, "primary_topic": {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.39649999141693115, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.39649999141693115, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.3391000032424927, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11010", "display_name": "Logic, Reasoning, and Knowledge", "score": 0.030799999833106995, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.00732", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.00732", "pdf_url": "https://arxiv.org/pdf/2510.00732", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4407571385", "doi": "https://doi.org/10.48550/arxiv.2502.08908", "title": "Reinforced Large Language Model is a formal theorem prover", "publication_year": 2025, "publication_date": "2025-02-13", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5020904548", "display_name": "Zhiling Luo", "orcid": "https://orcid.org/0000-0002-0540-7307", "observed_orcids": ["https://orcid.org/0000-0002-0540-7307"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Luo, Zhiling", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"To": [0], "take": [1], "advantage": [2], "of": [3], "Large": [4], "Language": [5], "Model": [6], "in": [7], "theorem": [8], "formalization": [9], "and": [10, 29], "proof,": [11], "we": [12], "propose": [13], "a": [14, 45], "reinforcement": [15], "learning": [16], "framework": [17], "to": [18, 43], "iteratively": [19], "optimize": [20], "the": [21, 33], "pretrained": [22], "LLM": [23], "by": [24], "rolling": [25], "out": [26], "next": [27], "tactics": [28], "comparing": [30], "them": [31], "with": [32, 49], "expected": [34], "ones.": [35], "The": [36], "experiment": [37], "results": [38], "show": [39], "that": [40], "it": [41], "helps": [42], "achieve": [44], "higher": [46], "accuracy": [47], "compared": [48], "directly": [50], "fine-tuned": [51], "LLM.": [52]}, "primary_topic": {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.4253000020980835, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.4253000020980835, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.23749999701976776, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.05869999900460243, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2502.08908", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2502.08908", "pdf_url": "https://arxiv.org/pdf/2502.08908", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417119478", "doi": "https://doi.org/10.48550/arxiv.2506.06607", "title": "Training-Free Tokenizer Transplantation via Orthogonal Matching Pursuit", "publication_year": 2025, "publication_date": "2025-06-07", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5069114386", "display_name": "Charles Goddard", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Goddard, Charles", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5033878635", "display_name": "Fernando Fernandes Neto", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Neto, Fernando Fernandes", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"We": [0, 172], "present": [1], "a": [2, 31, 53, 140], "training-free": [3], "method": [4, 175], "to": [5, 105], "transplant": [6], "tokenizers": [7], "in": [8, 38, 47], "pretrained": [9, 154], "large": [10, 125], "language": [11], "models": [12], "(LLMs)": [13], "by": [14], "reconstructing": [15], "unseen": [16], "token": [17, 29], "embeddings": [18], "via": [19], "Orthogonal": [20], "Matching": [21], "Pursuit": [22], "(OMP).": [23], "Specifically,": [24], "we": [25], "approximate": [26], "each": [27, 43], "out-of-vocabulary": [28], "as": [30, 139], "sparse": [32, 64], "linear": [33], "combination": [34], "of": [35, 56, 90, 153], "shared": [36, 57], "tokens,": [37, 59], "two": [39, 74], "phases:": [40], "first,": [41], "compute": [42], "new": [44, 158], "token's": [45], "representation": [46], "the": [48, 68, 91, 119, 177], "donor": [49], "embedding": [50, 71], "space": [51], "with": [52, 157], "small": [54], "dictionary": [55], "anchor": [58], "then": [60], "transfer": [61], "these": [62], "same": [63], "coefficients": [65], "back": [66], "into": [67, 176], "base": [69, 92], "model's": [70, 93], "space.": [72], "On": [73], "challenging": [75], "cross-tokenizer": [76, 161], "tasks--Llama$\\to$Mistral": [77], "NeMo": [78], "(12B)": [79], "and": [80, 109, 168], "Qwen$\\to$Llama": [81], "(1B)--we": [82], "show": [83], "that": [84], "OMP": [85, 116], "achieves": [86, 118], "best": [87, 120], "zero-shot": [88, 100], "preservation": [89], "performance": [94], "across": [95], "multiple": [96], "benchmarks,": [97], "while": [98], "other": [99], "approaches": [101, 111], "degrade": [102], "significantly.": [103], "Compared": [104], "baselines": [106], "(zero-init,": [107], "mean-init,": [108], "existing": [110], "like": [112], "WECHSEL,": [113], "FOCUS,": [114], "ZETT),": [115], "consistently": [117], "overall": [121], "performance,": [122], "effectively": [123], "bridging": [124], "tokenizer": [126], "discrepancies": [127], "without": [128], "gradient": [129], "updates.": [130], "Our": [131], "analysis": [132], "further": [133], "identifies": [134], "mismatched": [135], "numerical": [136], "tokenization": [137], "schemes": [138], "critical": [141], "challenge": [142], "for": [143, 181], "preserving": [144], "mathematical": [145], "reasoning": [146], "capabilities.": [147], "This": [148], "technique": [149], "enables": [150], "direct": [151], "reuse": [152], "model": [155], "weights": [156], "tokenizers,": [159], "facilitating": [160], "knowledge": [162], "distillation,": [163], "speculative": [164], "decoding,": [165], "ensembling,": [166], "merging,": [167], "domain-specific": [169], "vocabulary": [170, 184], "adaptations.": [171], "integrate": [173], "our": [174], "open-source": [178], "mergekit-tokensurgeon": [179], "tool": [180], "post": [182], "hoc": [183], "realignment.": [185]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.9262999892234802, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.9262999892234802, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.05869999900460243, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.0052999998442828655, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2506.06607", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2506.06607", "pdf_url": "https://arxiv.org/pdf/2506.06607", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7114790190", "doi": "https://doi.org/10.5445/ir/1000188274", "title": "Can Large Language Models Autoformalize Kinematics?", "publication_year": 2025, "publication_date": "2025-01-01", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Kabra, Aditi", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kabra, Aditi", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Laurent, Jonathan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Laurent, Jonathan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Bharadwaj, Sagar", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Bharadwaj, Sagar", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Martins, Ruben", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Martins, Ruben", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Mitsch, Stefan", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Mitsch, Stefan", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Platzer, Andr\u00e9", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Platzer, Andr\u00e9", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "Irfan, Ahmed", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Irfan, Ahmed", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": null, "display_name": "Kaufmann, Daniela", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Kaufmann, Daniela", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Autonomous": [0], "cyber-physical": [1, 43], "systems": [2], "like": [3], "robots": [4], "and": [5, 55, 104, 120, 127], "self-driving": [6], "cars": [7], "could": [8], "greatly": [9], "benefit": [10], "from": [11, 81, 177], "using": [12], "formal": [13, 38], "methods": [14], "to": [15, 31, 142, 180], "reason": [16], "reliably": [17], "about": [18], "their": [19], "control": [20], "decisions.": [21], "However,": [22], "before": [23], "a": [24, 37, 47, 57, 96, 107, 169, 181], "problem": [25, 75], "can": [26, 68], "be": [27, 32], "solved": [28], "it": [29], "needs": [30], "stated.": [33], "This": [34, 59, 167], "requires": [35, 52], "writing": [36], "physics": [39, 85, 145], "model": [40, 108, 115], "of": [41, 100, 150], "the": [42, 70, 91, 101, 136, 140, 143], "system,": [44], "which": [45], "is": [46, 78, 93, 116, 156], "complex": [48], "task": [49], "that": [50], "traditionally": [51], "human": [53], "expertise": [54], "becomes": [56], "bottleneck.": [58], "paper": [60], "experimentally": [61], "studies": [62], "whether": [63, 133], "Large": [64], "Language": [65], "Models": [66], "(LLMs)": [67], "automate": [69], "formalization": [71], "process.": [72], "A": [73, 147], "20": [74], "benchmark": [76], "suite": [77], "designed": [79], "drawing": [80], "undergrad-": [82], "uate": [83], "level": [84], "kinematics": [86], "problems.": [87], "In": [88], "each": [89], "problem,": [90], "LLM": [92], "provided": [94], "with": [95, 185], "natural": [97, 178], "language": [98, 179], "description": [99], "objects\u2019": [102], "motion": [103], "must": [105], "produce": [106], "in": [109], "differential": [110], "game": [111], "logic": [112, 184], "(dGL).": [113], "The": [114], "(1)": [117], "syntax": [118], "checked": [119], "iteratively": [121], "refined": [122], "based": [123], "on": [124], "parser": [125], "feedback,": [126], "(2)": [128], "semantically": [129], "evaluated": [130], "by": [131], "checking": [132], "symbolically": [134], "executing": [135], "dGL": [137], "formula": [138], "recovers": [139], "solution": [141], "original": [144], "problem.": [146], "success": [148], "rate": [149], "70%": [151], "(best": [152], "over": [153], "5": [154], "samples)": [155], "achieved.": [157], "We": [158], "analyze": [159], "failing": [160], "cases,": [161], "identifying": [162], "directions": [163], "for": [164, 173], "future": [165], "improvement.": [166], "provides": [168], "first": [170], "quantitative": [171], "baseline": [172], "LLM-based": [174], "autofor-": [175], "malization": [176], "hybrid": [182], "games": [183], "continuous": [186], "dynamics.": [187]}, "primary_topic": {"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.43970000743865967, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.43970000743865967, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.1680999994277954, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11010", "display_name": "Logic, Reasoning, and Knowledge", "score": 0.11819999665021896, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5445/ir/1000188274", "is_oa": true, "landing_page_url": "https://doi.org/10.5445/ir/1000188274", "pdf_url": null, "source": {"id": "https://openalex.org/S7407052948", "display_name": "KITopen", "issn_l": null, "issn": null, "is_oa": false, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I102335020", "host_organization_name": "Karlsruhe Institute of Technology", "host_organization_lineage": ["https://openalex.org/I102335020"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "article-journal"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7125367920", "doi": "https://doi.org/10.16926/pe.2025.18.12", "title": "Education in the Age of Artificial Intelligence. Conclusions from Sweden\u2019s Decision for Pedagogy and the Design of Educational Environments", "publication_year": 2025, "publication_date": "2025-01-01", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5027845227", "display_name": "Izabela Zi\u0119bacz", "orcid": "https://orcid.org/0000-0002-4275-0550", "observed_orcids": ["https://orcid.org/0000-0002-4275-0550"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Izabela Zi\u0119bacz", "raw_affiliation_strings": [], "raw_orcid": "https://orcid.org/0000-0002-4275-0550", "affiliations": []}], "abstract_inverted_index": {"The": [0, 34, 116, 139], "decision": [1], "of": [2, 26, 31, 41, 47, 72, 84, 129], "the": [3, 24, 29, 39, 82, 127, 142, 155], "Swedish": [4], "government": [5], "to": [6, 76, 122], "limit": [7], "digital": [8, 42], "technologies": [9], "in": [10, 28, 38, 80, 94, 100, 107, 150], "early": [11], "school": [12], "education": [13], "has": [14], "been": [15], "taken": [16], "as": [17], "a": [18, 133], "starting": [19], "point": [20], "for": [21, 144], "reflection": [22], "on": [23], "design": [25], "didactics": [27], "age": [30], "artificial": [32, 158], "intelligence.": [33], "analysis": [35], "was": [36], "embedded": [37], "framework": [40], "pedagogy,": [43], "Marian": [44], "Mazur\u2019s": [45], "theory": [46], "autonomous": [48, 164], "systems,": [49], "and": [50, 57, 65, 102, 105, 157, 163], "pedagogical": [51], "cybernetics.": [52], "OECD": [53], "data": [54], "(PIAAC,": [55], "PISA)": [56], "national": [58], "statistics": [59], "(GUS,": [60], "BN)": [61], "concerning": [62], "basic": [63], "skills": [64, 89], "reading": [66, 101], "practices": [67, 153], "were": [68], "used.": [69], "A": [70], "comparison": [71], "trends": [73], "from": [74], "2011/12": [75], "2022/23": [77], "shows": [78], "that": [79, 119], "Sweden": [81], "percentage": [83], "people": [85], "with": [86, 132], "very": [87], "low": [88], "remains": [90], "stable": [91], "(12\u201314%),": [92], "while": [93], "Poland": [95], "it": [96], "reaches": [97], "almost": [98], "40%": [99], "mathematical": [103], "reasoning": [104], "48%": [106], "problem": [108], "solving": [109], "(see": [110], "also": [111], "D\u00edaz": [112], "et": [113], "al.,": [114], "2024).": [115], "results": [117], "indicate": [118], "mere": [120], "access": [121], "technology": [123], "does": [124], "not": [125], "guarantee": [126], "development": [128], "competences": [130], "and,": [131], "weaker": [134], "foundation,": [135], "may": [136], "deepen": [137], "inequalities.": [138], "conclusions": [140], "emphasize": [141], "need": [143], "sequential": [145], "didactic": [146], "models": [147], "\u201coffline-first,": [148], "AI-assisted,\u201d": [149], "which": [151], "traditional": [152], "form": [154], "foundation": [156], "intelligence": [159], "supports": [160], "only": [161], "critical": [162], "thinking.": [165]}, "primary_topic": {"id": "https://openalex.org/T13081", "display_name": "Digital Education and Society", "score": 0.8755000233650208, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T13081", "display_name": "Digital Education and Society", "score": 0.8755000233650208, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T14414", "display_name": "Artificial Intelligence in Education", "score": 0.10369999706745148, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11300", "display_name": "Global Educational Policies and Reforms", "score": 0.002400000113993883, "subfield": {"id": "https://openalex.org/subfields/3320", "display_name": "Political Science and International Relations"}, "field": {"id": "https://openalex.org/fields/33", "display_name": "Social Sciences"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}], "primary_location": {"id": "doi:10.16926/pe.2025.18.12", "is_oa": true, "landing_page_url": "https://doi.org/10.16926/pe.2025.18.12", "pdf_url": null, "source": {"id": "https://openalex.org/S4210199347", "display_name": "Podstawy Edukacji", "issn_l": "2081-2264", "issn": ["2081-2264", "2658-2848"], "is_oa": true, "is_in_doaj": true, "is_core": false, "listed_in": ["doaj", "erih-plus"], "host_organization": "https://openalex.org/P4310317371", "host_organization_name": "Jan D\u0142ugosz University", "host_organization_lineage": ["https://openalex.org/P4310317371"], "host_organization_lineage_names": ["Jan D\u0142ugosz University"], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Podstawy Edukacji", "raw_type": "journal-article"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4414975431", "doi": "https://doi.org/10.48550/arxiv.2510.05283", "title": "Beyond Monolithic Rewards: A Hybrid and Multi-Aspect Reward Optimization for MLLM Alignment", "publication_year": 2025, "publication_date": "2025-10-06", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5104772578", "display_name": "Radha Gulhane", "orcid": "https://orcid.org/0009-0009-2591-1082", "observed_orcids": ["https://orcid.org/0009-0009-2591-1082"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Gulhane, Radha", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5125969895", "display_name": "Sathish Reddy Indurthi", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Indurthi, Sathish Reddy", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Aligning": [0], "multimodal": [1, 136], "large": [2], "language": [3], "models": [4], "(MLLMs)": [5], "with": [6, 87], "human": [7, 32, 74], "preferences": [8], "often": [9, 19], "relies": [10], "on": [11, 168], "single-signal,": [12], "model-based": [13, 59], "reward": [14, 40, 50, 56, 64, 105, 143], "methods.": [15], "Such": [16], "monolithic": [17], "rewards": [18, 95], "lack": [20], "confidence": [21], "calibration": [22], "across": [23, 134, 160], "domain-specific": [24, 81], "tasks,": [25], "fail": [26], "to": [27, 96, 106, 121], "capture": [28], "diverse": [29], "aspects": [30], "of": [31, 158, 178], "preferences,": [33], "and": [34, 39, 73, 76, 100, 109, 118, 141, 162, 186], "require": [35], "extensive": [36], "data": [37], "annotation": [38], "model": [41, 65, 148, 172], "training.": [42], "In": [43], "this": [44], "work,": [45], "we": [46, 91], "propose": [47], "a": [48, 62, 102, 116, 174], "hybrid": [49, 140], "modeling": [51], "framework": [52, 114], "that": [53], "integrates": [54], "complementary": [55], "paradigms:": [57], "(i)": [58], "rewards,": [60, 79], "where": [61, 80], "learned": [63], "predicts": [66], "scalar": [67], "or": [68], "vector": [69], "scores": [70], "from": [71], "synthetic": [72], "feedback,": [75], "(ii)": [77], "rule-based": [78], "heuristics": [82], "provide": [83], "explicit": [84], "correctness": [85], "signals": [86], "confidence.": [88], "Beyond": [89], "accuracy,": [90], "further": [92], "incorporate": [93], "multi-aspect": [94, 142], "enforce": [97], "instruction": [98], "adherence": [99], "introduce": [101], "generalized": [103], "length-penalty": [104], "stabilize": [107], "training": [108], "improve": [110], "performance.": [111], "The": [112], "proposed": [113], "provides": [115], "flexible": [117], "effective": [119], "approach": [120], "aligning": [122], "MLLMs": [123], "through": [124], "reinforcement": [125], "learning": [126], "policy": [127], "optimization.": [128], "Our": [129, 145], "experiments": [130], "show": [131], "consistent": [132], "improvements": [133], "different": [135], "benchmarks": [137], "when": [138], "applying": [139], "modeling.": [144], "best": [146], "performing": [147], "in": [149, 183], "the": [150, 171], "3B": [151], "family": [152], "achieves": [153, 173], "an": [154], "overall": [155], "average": [156, 176], "improvement": [157, 177], "~9.5%": [159], "general": [161], "math": [163], "reasoning": [164, 185], "tasks.": [165], "Focusing": [166], "specifically": [167], "mathematical": [169, 184], "benchmarks,": [170], "significant": [175], "~16%,": [179], "highlighting": [180], "its": [181], "effectiveness": [182], "problem": [187], "solving.": [188]}, "primary_topic": {"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.951200008392334, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11714", "display_name": "Multimodal Machine Learning Applications", "score": 0.951200008392334, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11307", "display_name": "Domain Adaptation and Few-Shot Learning", "score": 0.010200000368058681, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.006000000052154064, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.05283", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.05283", "pdf_url": "https://arxiv.org/pdf/2510.05283", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4408074280", "doi": "https://doi.org/10.1142/s0218202525500228", "title": "Representation and regression problems in neural networks: Relaxation, generalization, and numerics", "publication_year": 2025, "publication_date": "2025-03-01", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Kang Liu", "orcid": "https://orcid.org/0000-0002-9486-578X", "observed_orcids": ["https://orcid.org/0000-0002-9486-578X"]}, "institutions": [{"id": "https://openalex.org/I1294671590", "display_name": "Centre National de la Recherche Scientifique", "ror": "https://ror.org/02feahw73", "country_code": "FR", "type": "government", "lineage": ["https://openalex.org/I1294671590"]}, {"id": "https://openalex.org/I4210112333", "display_name": "Institut de Math\u00e9matiques de Bourgogne", "ror": "https://ror.org/021f0sa24", "country_code": "FR", "type": "facility", "lineage": ["https://openalex.org/I1294671590", "https://openalex.org/I4210112333", "https://openalex.org/I4405253280"]}], "countries": ["FR"], "is_corresponding": false, "raw_author_name": "Kang Liu", "raw_affiliation_strings": ["Institut de Math\u00e9matiques de Bourgogne, Universit\u00e9 Bourgogne Europe, CNRS, Dijon 21000, France"], "raw_orcid": "https://orcid.org/0000-0002-9486-578X", "affiliations": [{"raw_affiliation_string": "Institut de Math\u00e9matiques de Bourgogne, Universit\u00e9 Bourgogne Europe, CNRS, Dijon 21000, France", "institution_ids": ["https://openalex.org/I1294671590", "https://openalex.org/I4210112333"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5088702266", "display_name": "Enrique Zuazua", "orcid": "https://orcid.org/0000-0002-1377-0958", "observed_orcids": ["https://orcid.org/0000-0002-1377-0958"]}, "institutions": [{"id": "https://openalex.org/I136040515", "display_name": "Universidad de Deusto", "ror": "https://ror.org/00ne6sr39", "country_code": "ES", "type": "education", "lineage": ["https://openalex.org/I136040515"]}, {"id": "https://openalex.org/I181369854", "display_name": "Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg", "ror": "https://ror.org/00f7hpc57", "country_code": "DE", "type": "education", "lineage": ["https://openalex.org/I181369854"]}, {"id": "https://openalex.org/I63634437", "display_name": "Universidad Aut\u00f3noma de Madrid", "ror": "https://ror.org/01cby8j38", "country_code": "ES", "type": "education", "lineage": ["https://openalex.org/I63634437"]}], "countries": ["DE", "ES"], "is_corresponding": false, "raw_author_name": "Enrique Zuazua", "raw_affiliation_strings": ["Departamento de Matem\u00e1ticas, Universidad Aut\u00f3noma de Madrid, Madrid 28049, Spain", "Department of Mathematics, Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg, Erlangen 91058, Germany", "Fundaci\u00f3n Deusto, Bilbao 48007, Basque Country, Spain"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Departamento de Matem\u00e1ticas, Universidad Aut\u00f3noma de Madrid, Madrid 28049, Spain", "institution_ids": ["https://openalex.org/I63634437"]}, {"raw_affiliation_string": "Department of Mathematics, Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg, Erlangen 91058, Germany", "institution_ids": ["https://openalex.org/I181369854"]}, {"raw_affiliation_string": "Fundaci\u00f3n Deusto, Bilbao 48007, Basque Country, Spain", "institution_ids": ["https://openalex.org/I136040515"]}]}], "abstract_inverted_index": {"In": [0], "this": [1], "work,": [2], "we": [3, 33, 81, 109], "address": [4], "three": [5], "non-convex": [6], "optimization": [7], "problems": [8, 36, 88, 98], "associated": [9], "with": [10, 116], "the": [11, 43, 53, 65, 78, 83, 86, 103, 127], "training": [12], "of": [13, 45, 67, 85], "shallow": [14, 121], "neural": [15], "networks": [16], "(NNs)": [17], "for": [18, 26, 52, 119], "exact": [19], "and": [20, 73, 89], "approximate": [21], "representation,": [22], "as": [23, 25], "well": [24], "regression": [27], "tasks.": [28], "Through": [29], "a": [30, 39, 111], "mean-field": [31], "approach,": [32], "convexify": [34], "these": [35, 71, 96], "and,": [37], "applying": [38], "representer": [40], "theorem,": [41], "prove": [42], "absence": [44], "relaxation": [46], "gaps.": [47], "We": [48], "establish": [49], "generalization": [50], "bounds": [51], "resulting": [54], "NN": [55], "solutions,": [56], "assessing": [57], "their": [58], "predictive": [59], "performance": [60], "on": [61, 70], "test": [62], "datasets,": [63, 95, 108], "analyzing": [64], "impact": [66], "key": [68], "hyperparameters": [69], "bounds,": [72], "proposing": [74], "optimal": [75], "choices.": [76], "On": [77], "computational": [79], "side,": [80], "examine": [82], "discretization": [84], "convexified": [87], "derive": [90], "convergence": [91], "rates.": [92], "For": [93, 106], "low-dimensional": [94], "discretized": [97], "are": [99], "efficiently": [100], "solvable": [101], "using": [102], "simplex": [104], "method.": [105], "high-dimensional": [107], "propose": [110], "sparsification": [112], "algorithm": [113], "that,": [114], "combined": [115], "gradient": [117], "descent": [118], "over-parameterized": [120], "NNs,": [122], "yields": [123], "effective": [124], "solutions": [125], "to": [126], "primal": [128], "problems.": [129]}, "primary_topic": {"id": "https://openalex.org/T10320", "display_name": "Neural Networks and Applications", "score": 0.5156000256538391, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10320", "display_name": "Neural Networks and Applications", "score": 0.5156000256538391, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11612", "display_name": "Stochastic Gradient Optimization Techniques", "score": 0.46950000524520874, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12676", "display_name": "Machine Learning and ELM", "score": 0.0035000001080334187, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.1142/s0218202525500228", "is_oa": false, "landing_page_url": "https://doi.org/10.1142/s0218202525500228", "pdf_url": null, "source": {"id": "https://openalex.org/S24332595", "display_name": "Mathematical Models and Methods in Applied Sciences", "issn_l": "0218-2025", "issn": ["0218-2025", "1793-4060", "1793-6314"], "is_oa": false, "is_in_doaj": false, "is_core": true, "listed_in": ["cwts-core", "jufo-2", "norway-2"], "host_organization": "https://openalex.org/P4310319815", "host_organization_name": "World Scientific", "host_organization_lineage": ["https://openalex.org/P4310319815"], "host_organization_lineage_names": ["World Scientific"], "type": "journal"}, "license": null, "license_id": null, "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Mathematical Models and Methods in Applied Sciences", "raw_type": "journal-article"}, "cited_by_count": 2, "is_retracted": false}, {"id": "https://openalex.org/W4416785351", "doi": "https://doi.org/10.48550/arxiv.2511.21104", "title": "BRIDGE: Building Representations In Domain Guided Program Synthesis", "publication_year": 2025, "publication_date": "2025-11-26", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5061177581", "display_name": "Robert Joseph George", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I122411786", "display_name": "California Institute of Technology", "ror": "https://ror.org/05dxps055", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I122411786"]}, {"id": "https://openalex.org/I1311688040", "display_name": "Amazon (United States)", "ror": "https://ror.org/04mv4n011", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I1311688040"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "George, Robert Joseph", "raw_affiliation_strings": ["Amazon", "California Institute of Technology"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Amazon", "institution_ids": ["https://openalex.org/I1311688040"]}, {"raw_affiliation_string": "California Institute of Technology", "institution_ids": ["https://openalex.org/I122411786"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5053914366", "display_name": "Carson Eisenach", "orcid": "https://orcid.org/0000-0002-5457-3328", "observed_orcids": ["https://orcid.org/0000-0002-5457-3328"]}, "institutions": [{"id": "https://openalex.org/I1311688040", "display_name": "Amazon (United States)", "ror": "https://ror.org/04mv4n011", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I1311688040"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Eisenach, Carson", "raw_affiliation_strings": ["Amazon"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Amazon", "institution_ids": ["https://openalex.org/I1311688040"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5017544719", "display_name": "Udaya Ghai", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I1311688040", "display_name": "Amazon (United States)", "ror": "https://ror.org/04mv4n011", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I1311688040"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Ghai, Udaya", "raw_affiliation_strings": ["Amazon"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Amazon", "institution_ids": ["https://openalex.org/I1311688040"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5004130134", "display_name": "Dominique Perrault-Joncas", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Perrault-Joncas, Dominique", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5014498545", "display_name": "Anima Anandkumar", "orcid": "https://orcid.org/0000-0002-6974-6797", "observed_orcids": ["https://orcid.org/0000-0002-6974-6797"]}, "institutions": [{"id": "https://openalex.org/I122411786", "display_name": "California Institute of Technology", "ror": "https://ror.org/05dxps055", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I122411786"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Anandkumar, Anima", "raw_affiliation_strings": ["California Institute of Technology"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "California Institute of Technology", "institution_ids": ["https://openalex.org/I122411786"]}]}, {"author_position": "last", "author": {"id": null, "display_name": "Foster, Dean", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I1311688040", "display_name": "Amazon (United States)", "ror": "https://ror.org/04mv4n011", "country_code": "US", "type": "company", "lineage": ["https://openalex.org/I1311688040"]}], "countries": ["US"], "is_corresponding": false, "raw_author_name": "Foster, Dean", "raw_affiliation_strings": ["Amazon"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Amazon", "institution_ids": ["https://openalex.org/I1311688040"]}]}], "abstract_inverted_index": {"Large": [0], "language": [1], "models": [2], "can": [3, 125], "generate": [4], "plausible": [5], "code,": [6, 35], "but": [7], "remain": [8], "brittle": [9], "for": [10, 57, 96, 186], "formal": [11, 36, 200], "verification": [12], "in": [13], "proof": [14, 41, 201], "assistants": [15], "such": [16], "as": [17, 92], "Lean.": [18], "A": [19], "central": [20], "scalability": [21], "challenge": [22], "is": [23], "that": [24, 139, 172], "verified": [25, 188], "synthesis": [26, 189], "requires": [27], "consistent": [28], "artifacts": [29, 48], "across": [30], "several": [31], "coupled": [32], "domains:": [33, 67], "executable": [34, 114, 196], "specifications,": [37], "theorem": [38, 99], "statements,": [39], "and": [40, 70, 72, 101, 108, 124, 198], "attempts.": [42], "Existing": [43], "approaches": [44], "often": [45, 83], "treat": [46], "these": [47, 173], "separately.": [49], "We": [50, 136], "present": [51], "BRIDGE,": [52], "a": [53, 85, 93, 177, 183], "structured": [54], "prompting": [55, 123, 141], "framework": [56, 185], "multi-artifact": [58], "program": [59], "synthesis.": [60], "BRIDGE": [61, 82, 111, 181], "decomposes": [62], "generation": [63, 134], "into": [64], "three": [65], "interconnected": [66], "Code,": [68], "Specification,": [69], "Theorem/Proof,": [71], "uses": [73], "domain-specific": [74], "intermediate": [75, 174], "reasoning": [76, 159], "to": [77, 118, 148], "connect": [78], "them.": [79], "In": [80], "Lean,": [81], "follows": [84], "code-first": [86], "workflow,": [87], "using": [88], "the": [89, 192], "generated": [90], "implementation": [91], "semantic": [94], "anchor": [95], "downstream": [97], "specification,": [98], "statement,": [100], "proof-attempt": [102], "generation.": [103, 202], "Across": [104], "178": [105], "algorithmic": [106], "problems": [107], "five": [109], "LLMs,": [110], "improves": [112, 142], "Lean": [113, 165], "correctness": [115, 197], "by": [116, 146], "up": [117, 147], "nearly": [119, 162], "1.5x": [120, 163], "over": [121], "direct": [122], "be": [126], "roughly": [127], "2x": [128], "more": [129], "sample": [130], "efficient": [131], "at": [132], "comparable": [133], "lengths.": [135], "further": [137], "find": [138], "specification-oriented": [140], "Python": [143], "pass": [144, 166], "rates": [145], "17.5": [149], "percentage": [150], "points.": [151], "Beyond": [152], "inference-time": [153], "prompting,": [154], "supervised": [155], "fine-tuning": [156], "on": [157], "BRIDGE-style": [158], "traces": [160], "yields": [161], "higher": [164], "success": [167], "than": [168], "code-only": [169], "fine-tuning,": [170], "suggesting": [171], "representations": [175], "provide": [176], "learnable": [178], "inductive": [179], "bias.": [180], "provides": [182], "practical": [184], "scaling": [187], "while": [190], "highlighting": [191], "remaining": [193], "gap": [194], "between": [195], "full": [199]}, "primary_topic": {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.5253999829292297, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.5253999829292297, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10533", "display_name": "Teaching and Learning Programming", "score": 0.14000000059604645, "subfield": {"id": "https://openalex.org/subfields/1706", "display_name": "Computer Science Applications"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10260", "display_name": "Software Engineering Research", "score": 0.09449999779462814, "subfield": {"id": "https://openalex.org/subfields/1710", "display_name": "Information Systems"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2511.21104", "is_oa": true, "landing_page_url": "https://arxiv.org/abs/2511.21104", "pdf_url": "https://arxiv.org/pdf/2511.21104", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7109081556", "doi": "https://doi.org/10.5281/zenodo.17827520", "title": "Higher-Order Reasoning through Iterative Self-Refinement", "publication_year": 2025, "publication_date": "2025-12-05", "type": "article", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "Revista, Zen", "orcid": "https://orcid.org/0009-0007-6299-2008", "observed_orcids": ["https://orcid.org/0009-0007-6299-2008"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Revista, Zen", "raw_affiliation_strings": [], "raw_orcid": "https://orcid.org/0009-0007-6299-2008", "affiliations": []}, {"author_position": "last", "author": {"id": null, "display_name": "IA, 10", "orcid": "https://orcid.org/0009-0007-6299-2008", "observed_orcids": ["https://orcid.org/0009-0007-6299-2008"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "IA, 10", "raw_affiliation_strings": [], "raw_orcid": "https://orcid.org/0009-0007-6299-2008", "affiliations": []}], "abstract_inverted_index": {"Higher-order": [0], "reasoning,": [1, 142, 180], "characterized": [2], "by": [3, 156], "complex": [4], "logical": [5], "deductions,": [6], "multi-step": [7], "problem-solving,": [8], "and": [9, 34, 90, 101, 119, 133, 144, 147, 161, 175, 185], "abstract": [10], "conceptual": [11], "understanding,": [12, 146], "remains": [13], "a": [14, 63, 127, 171], "formidable": [15], "challenge": [16], "for": [17, 50], "artificial": [18], "intelligence": [19], "systems,": [20], "particularly": [21], "large": [22], "language": [23, 145], "models": [24], "(LLMs).": [25], "While": [26], "LLMs": [27, 94], "demonstrate": [28], "remarkable": [29], "capabilities": [30], "in": [31], "generating": [32], "coherent": [33], "contextually": [35], "relevant": [36], "text,": [37], "their": [38, 76, 98], "initial": [39, 86], "outputs": [40], "often": [41], "contain": [42], "flaws,": [43], "inconsistencies,": [44], "or": [45], "lack": [46], "the": [47, 57], "depth": [48], "required": [49], "true": [51], "higher-order": [52, 177], "tasks.": [53], "This": [54], "paper": [55], "explores": [56], "paradigm": [58], "of": [59, 130, 179], "iterative": [60], "self-refinement": [61, 92], "as": [62, 97, 140], "potent": [64], "mechanism": [65], "to": [66, 73, 95, 108, 182], "bridge": [67], "this": [68], "gap,": [69], "enabling": [70], "AI": [71, 167], "systems": [72, 168], "progressively": [74], "enhance": [75], "reasoning": [77, 116], "capabilities.": [78], "Drawing": [79], "inspiration": [80], "from": [81], "human": [82], "cognitive": [83], "processes": [84], "where": [85], "thoughts": [87], "are": [88], "critiqued": [89], "refined,": [91], "empowers": [93], "act": [96], "own": [99], "critics": [100], "optimizers.": [102], "We": [103], "delve": [104], "into": [105], "various": [106], "approaches": [107], "self-refinement,": [109], "including": [110], "internal": [111], "feedback": [112], "generation,": [113], "progressive": [114], "back-translation,": [115], "topology": [117], "integration,": [118], "gap-driven": [120], "enhancement,": [121], "demonstrating": [122], "how": [123], "these": [124], "techniques": [125], "facilitate": [126], "virtuous": [128], "cycle": [129], "critique,": [131], "revision,": [132], "performance": [134], "improvement": [135], "across": [136], "diverse": [137], "domains": [138], "such": [139], "mathematical": [141], "visual": [143], "sequential": [148], "task": [149], "planning.": [150], "The": [151], "core": [152], "argument": [153], "is": [154], "that": [155], "moving": [157], "beyond": [158], "single-pass": [159], "generation": [160], "embracing": [162], "an": [163], "iterative,": [164], "feedback-driven": [165], "approach,": [166], "can": [169], "achieve": [170], "more": [172, 183], "robust,": [173], "accurate,": [174], "truly": [176], "form": [178], "leading": [181], "reliable": [184], "sophisticated": [186], "intelligent": [187], "agents.": [188]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.6596999764442444, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.6596999764442444, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.09369999915361404, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11010", "display_name": "Logic, Reasoning, and Knowledge", "score": 0.0697999969124794, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5281/zenodo.17827520", "is_oa": true, "landing_page_url": "https://doi.org/10.5281/zenodo.17827520", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400562", "display_name": "Zenodo (CERN European Organization for Nuclear Research)", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I67311998", "host_organization_name": "European Organization for Nuclear Research", "host_organization_lineage": ["https://openalex.org/I67311998"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "JournalArticle"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7103878345", "doi": "https://doi.org/10.5281/zenodo.17518042", "title": "Neuro-Symbolic Synthesis of Formal Proofs", "publication_year": 2025, "publication_date": "2025-11-23", "type": "article", "language": null, "authorships": [{"author_position": "first", "author": {"id": null, "display_name": "S\u00c9RGIO DE ANDRADE, PAULO", "orcid": "https://orcid.org/0009-0004-2555-3178", "observed_orcids": ["https://orcid.org/0009-0004-2555-3178"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "S\u00c9RGIO DE ANDRADE, PAULO", "raw_affiliation_strings": [], "raw_orcid": "https://orcid.org/0009-0004-2555-3178", "affiliations": []}], "abstract_inverted_index": {"This": [0, 63, 98, 149], "paper": [1], "investigates": [2], "the": [3, 10, 31, 55, 71, 88, 132, 139, 158, 173, 180, 189], "synthesis": [4, 160], "of": [5, 12, 20, 73, 176, 183, 191], "formal": [6, 61, 196], "mathematical": [7], "proofs": [8], "through": [9, 146], "integration": [11], "neural": [13, 43, 77, 140, 153, 177], "and": [14, 30, 50, 95, 104, 156, 200], "symbolic": [15, 22, 108, 133, 184], "AI": [16], "paradigms.": [17], "The": [18, 167], "limitations": [19], "purely": [21], "automated": [23, 192], "theorem": [24, 193], "provers,": [25], "such": [26, 110], "as": [27, 111], "combinatorial": [28], "explosion": [29], "need": [32], "for": [33, 60], "handcrafted": [34], "heuristics,": [35], "have": [36], "historically": [37], "hindered": [38], "their": [39], "efficacy.": [40], "Conversely,": [41], "large-scale": [42], "models,": [44], "while": [45], "proficient": [46], "at": [47, 123], "pattern": [48], "recognition": [49], "generating": [51], "human-like": [52], "text,": [53], "lack": [54], "rigorous": [56], "logical": [57, 121], "consistency": [58], "required": [59], "verification.": [62], "work": [64], "proposes": [65], "a": [66, 80, 107, 112, 128, 163], "neuro-symbolic": [67], "framework": [68], "that": [69, 171], "leverages": [70], "strengths": [72], "both": [74], "approaches.": [75], "A": [76], "component,": [78], "typically": [79], "transformer-based": [81], "language": [82], "model,": [83], "is": [84, 101], "employed": [85], "to": [86, 142], "guide": [87], "proof": [89, 113, 159], "search,": [90], "suggesting": [91], "promising": [92], "axioms,": [93], "tactics,": [94], "intermediate": [96], "lemmas.": [97], "intuition-driven": [99], "guidance": [100], "then": [102], "validated": [103], "executed": [105], "by": [106], "engine,": [109], "assistant": [114], "(e.g.,": [115], "Lean,": [116], "Coq,": [117], "Isabelle/HOL),": [118], "which": [119], "guarantees": [120], "soundness": [122], "every": [124], "step.": [125], "We": [126], "explore": [127], "verifier-in-the-loop": [129], "methodology": [130], "where": [131], "system": [134], "provides": [135], "structured": [136], "feedback,": [137], "enabling": [138], "model": [141, 154], "refine": [143], "its": [144], "heuristics": [145], "reinforcement": [147], "learning.": [148], "synergistic": [150], "loop": [151], "mitigates": [152], "hallucination": [155], "grounds": [157], "process": [161], "in": [162], "formal,": [164], "verifiable": [165], "context.": [166], "proposed": [168], "architecture": [169], "demonstrates": [170], "combining": [172], "statistical": [174], "inference": [175], "networks": [178], "with": [179], "deductive": [181], "power": [182], "reasoners": [185], "can": [186], "significantly": [187], "advance": [188], "state": [190], "proving,": [194], "making": [195], "methods": [197], "more": [198], "scalable": [199], "accessible.": [201]}, "primary_topic": {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.8398000001907349, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.8398000001907349, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.05260000005364418, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.0430000014603138, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.5281/zenodo.17518042", "is_oa": true, "landing_page_url": "https://doi.org/10.5281/zenodo.17518042", "pdf_url": null, "source": {"id": "https://openalex.org/S4306400562", "display_name": "Zenodo (CERN European Organization for Nuclear Research)", "issn_l": null, "issn": null, "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I67311998", "host_organization_name": "European Organization for Nuclear Research", "host_organization_lineage": ["https://openalex.org/I67311998"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": null, "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "JournalArticle"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W7116869808", "doi": "https://doi.org/10.48550/arxiv.2512.17079", "title": "Can Large Reasoning Models Improve Accuracy on Mathematical Tasks Using Flawed Thinking?", "publication_year": 2025, "publication_date": "2025-12-18", "type": "preprint", "language": null, "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5121043535", "display_name": "Saraswathy Amjith", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Amjith, Saraswathy", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5120840111", "display_name": "Mihika Dusad", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Dusad, Mihika", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5121089456", "display_name": "Neha Muramalla", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Muramalla, Neha", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5121080964", "display_name": "Shweta Shah", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Shah, Shweta", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Chain-of-thought": [0], "(CoT)": [1], "prompting": [2], "has": [3], "become": [4], "central": [5], "to": [6, 17, 30, 47, 144, 172], "mathematical": [7, 190], "reasoning": [8, 42, 83, 121, 152, 191], "in": [9, 192], "large": [10], "language": [11], "models,": [12], "yet": [13], "models": [14, 46], "remain": [15], "brittle": [16], "early": [18], "errors:": [19], "a": [20, 74, 82, 96, 185], "single": [21], "arithmetic": [22], "slip": [23], "or": [24, 81], "unjustified": [25, 87], "inference": [26], "typically": [27], "propagates": [28], "uncorrected": [29], "an": [31], "incorrect": [32], "final": [33], "answer.": [34], "We": [35], "investigate": [36], "whether": [37], "training": [38, 141, 150, 164, 176], "on": [39, 106, 116, 151], "intentionally": [40], "flawed": [41, 120, 173], "traces": [43, 174], "can": [44, 177], "teach": [45], "detect": [48], "and": [49, 90], "recover": [50], "from": [51, 62], "such": [52], "errors": [53, 153, 160], "without": [54, 181], "degrading": [55], "standard": [56, 104], "problem-solving": [57], "ability.": [58], "Using": [59], "competition-level": [60], "problems": [61, 108, 117], "MATH-lighteval,": [63], "we": [64], "generate": [65], "CoT": [66], "prefixes": [67], "containing": [68], "exactly": [69], "one": [70], "controlled": [71], "error,": [72], "either": [73], "calculation": [75, 159], "error": [76, 84, 148], "(sign": [77], "flips,": [78], "dropped": [79], "terms)": [80], "(misapplied": [85], "rules,": [86], "logical": [88], "steps),": [89], "fine-tune": [91], "Qwen3-4B": [92], "with": [93, 119, 162], "GRPO": [94], "using": [95], "binary": [97], "final-answer": [98], "reward.": [99], "Our": [100], "Mixed-CoT-RL": [101], "model": [102], "matches": [103], "RL": [105, 127], "clean": [107], "(41%": [109], "vs": [110, 123], "41%)": [111], "while": [112], "substantially": [113], "outperforming": [114], "it": [115], "prefilled": [118], "(24%": [122], "19%).": [124], "Notably,": [125], "clean-only": [126], "fine-tuning": [128], "degrades": [129], "robustness": [130, 156], "below": [131], "the": [132], "untuned": [133], "baseline": [134], "19%": [135], "vs.": [136], "20%),": [137], "indicating": [138], "that": [139, 170], "conventional": [140], "increases": [142], "susceptibility": [143], "misleading": [145], "prefills.": [146], "Among": [147], "types,": [149], "yields": [154], "greater": [155], "gains": [157], "than": [158], "alone,": [161], "mixed": [163], "performing": [165], "best.": [166], "These": [167], "findings": [168], "demonstrate": [169], "exposure": [171], "during": [175], "improve": [178], "error-recovery": [179], "behavior": [180], "sacrificing": [182], "accuracy,": [183], "suggesting": [184], "path": [186], "toward": [187], "more": [188], "robust": [189], "LLMs.": [193]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5679000020027161, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.5679000020027161, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T12522", "display_name": "Mathematics Education and Pedagogy", "score": 0.13279999792575836, "subfield": {"id": "https://openalex.org/subfields/2604", "display_name": "Applied Mathematics"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.043699998408555984, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2512.17079", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2512.17079", "pdf_url": "https://arxiv.org/pdf/2512.17079", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4408851327", "doi": "https://doi.org/10.1016/j.compbiomed.2025.109985", "title": "Synergistic eigenanalysis of covariance and Hessian matrices for enhanced binary classification on health datasets", "publication_year": 2025, "publication_date": "2025-03-25", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5001586722", "display_name": "Agus Hartoyo", "orcid": "https://orcid.org/0000-0002-8285-8689", "observed_orcids": ["https://orcid.org/0000-0002-8285-8689"]}, "institutions": [{"id": "https://openalex.org/I4405252653", "display_name": "Sano \u2013 Centre for Computational Personalised Medicine International Research Foundation", "ror": "https://ror.org/04h58p752", "country_code": null, "type": "facility", "lineage": ["https://openalex.org/I4405252653"]}, {"id": "https://openalex.org/I862893732", "display_name": "Telkom University", "ror": "https://ror.org/0004wsx81", "country_code": "ID", "type": "education", "lineage": ["https://openalex.org/I862893732"]}], "countries": ["ID"], "is_corresponding": true, "raw_author_name": "Agus Hartoyo", "raw_affiliation_strings": ["Sano - Centre for Computational Personalised Medicine, International Research Foundation, Krakow, Poland; School of Computing, Telkom University, Bandung, Indonesia. Electronic address: a.hartoyo@sanoscience.org"], "raw_orcid": "https://orcid.org/0000-0002-8285-8689", "affiliations": [{"raw_affiliation_string": "Sano - Centre for Computational Personalised Medicine, International Research Foundation, Krakow, Poland; School of Computing, Telkom University, Bandung, Indonesia. Electronic address: a.hartoyo@sanoscience.org", "institution_ids": ["https://openalex.org/I4405252653", "https://openalex.org/I862893732"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5037310658", "display_name": "Jan K. Argasi\u0144ski", "orcid": "https://orcid.org/0000-0002-2992-718X", "observed_orcids": ["https://orcid.org/0000-0002-2992-718X"]}, "institutions": [{"id": "https://openalex.org/I126596746", "display_name": "Jagiellonian University", "ror": "https://ror.org/03bqmcz70", "country_code": "PL", "type": "education", "lineage": ["https://openalex.org/I126596746"]}, {"id": "https://openalex.org/I4405252653", "display_name": "Sano \u2013 Centre for Computational Personalised Medicine International Research Foundation", "ror": "https://ror.org/04h58p752", "country_code": null, "type": "facility", "lineage": ["https://openalex.org/I4405252653"]}], "countries": ["PL"], "is_corresponding": false, "raw_author_name": "Jan Argasi\u0144ski", "raw_affiliation_strings": ["Sano - Centre for Computational Personalised Medicine, International Research Foundation, Krakow, Poland; Department of Human-Centered Artificial Intelligence, Institute of Applied Computer Science, Faculty of Physics, Astronomy and Applied Computer Science, Jagiellonian University, Krakow, Poland"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Sano - Centre for Computational Personalised Medicine, International Research Foundation, Krakow, Poland; Department of Human-Centered Artificial Intelligence, Institute of Applied Computer Science, Faculty of Physics, Astronomy and Applied Computer Science, Jagiellonian University, Krakow, Poland", "institution_ids": ["https://openalex.org/I126596746", "https://openalex.org/I4405252653"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5043605500", "display_name": "Aleksandra Trenk", "orcid": "https://orcid.org/0000-0002-5888-3550", "observed_orcids": ["https://orcid.org/0000-0002-5888-3550"]}, "institutions": [{"id": "https://openalex.org/I126596746", "display_name": "Jagiellonian University", "ror": "https://ror.org/03bqmcz70", "country_code": "PL", "type": "education", "lineage": ["https://openalex.org/I126596746"]}], "countries": ["PL"], "is_corresponding": false, "raw_author_name": "Aleksandra Trenk", "raw_affiliation_strings": ["Department of Neurophysiology and Chronobiology, Institute of Zoology and Biomedical Research, Faculty of Biology, Jagiellonian University, Krakow, Poland"], "raw_orcid": "https://orcid.org/0000-0002-5888-3550", "affiliations": [{"raw_affiliation_string": "Department of Neurophysiology and Chronobiology, Institute of Zoology and Biomedical Research, Faculty of Biology, Jagiellonian University, Krakow, Poland", "institution_ids": ["https://openalex.org/I126596746"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5092824136", "display_name": "Kinga Przybylska", "orcid": "https://orcid.org/0009-0003-9590-4909", "observed_orcids": ["https://orcid.org/0009-0003-9590-4909"]}, "institutions": [{"id": "https://openalex.org/I126596746", "display_name": "Jagiellonian University", "ror": "https://ror.org/03bqmcz70", "country_code": "PL", "type": "education", "lineage": ["https://openalex.org/I126596746"]}], "countries": ["PL"], "is_corresponding": false, "raw_author_name": "Kinga Przybylska", "raw_affiliation_strings": ["Department of Neurophysiology and Chronobiology, Institute of Zoology and Biomedical Research, Faculty of Biology, Jagiellonian University, Krakow, Poland; Doctoral School of Exact and Natural Sciences, Jagiellonian University, Krakow, Poland"], "raw_orcid": "https://orcid.org/0009-0003-9590-4909", "affiliations": [{"raw_affiliation_string": "Department of Neurophysiology and Chronobiology, Institute of Zoology and Biomedical Research, Faculty of Biology, Jagiellonian University, Krakow, Poland; Doctoral School of Exact and Natural Sciences, Jagiellonian University, Krakow, Poland", "institution_ids": ["https://openalex.org/I126596746"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5019815585", "display_name": "Anna B\u0142asiak", "orcid": "https://orcid.org/0000-0002-4017-2988", "observed_orcids": ["https://orcid.org/0000-0002-4017-2988"]}, "institutions": [{"id": "https://openalex.org/I126596746", "display_name": "Jagiellonian University", "ror": "https://ror.org/03bqmcz70", "country_code": "PL", "type": "education", "lineage": ["https://openalex.org/I126596746"]}], "countries": ["PL"], "is_corresponding": false, "raw_author_name": "Anna B\u0142asiak", "raw_affiliation_strings": ["Department of Neurophysiology and Chronobiology, Institute of Zoology and Biomedical Research, Faculty of Biology, Jagiellonian University, Krakow, Poland"], "raw_orcid": "https://orcid.org/0000-0002-4017-2988", "affiliations": [{"raw_affiliation_string": "Department of Neurophysiology and Chronobiology, Institute of Zoology and Biomedical Research, Faculty of Biology, Jagiellonian University, Krakow, Poland", "institution_ids": ["https://openalex.org/I126596746"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5047749964", "display_name": "Alessandro Crimi", "orcid": "https://orcid.org/0000-0001-5397-6363", "observed_orcids": ["https://orcid.org/0000-0001-5397-6363"]}, "institutions": [{"id": "https://openalex.org/I686019", "display_name": "AGH University of Krakow", "ror": "https://ror.org/00bas1c41", "country_code": "PL", "type": "education", "lineage": ["https://openalex.org/I686019"]}], "countries": ["PL"], "is_corresponding": false, "raw_author_name": "Alessandro Crimi", "raw_affiliation_strings": ["AGH University of Krakow, Krakow, Poland"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "AGH University of Krakow, Krakow, Poland", "institution_ids": ["https://openalex.org/I686019"]}]}], "abstract_inverted_index": {"Covariance": [0], "and": [1, 88, 118, 151, 159, 175, 180, 196, 254, 273, 285], "Hessian": [2, 50, 182, 255], "matrices": [3, 17, 256], "have": [4], "been": [5], "analyzed": [6], "separately": [7], "in": [8, 26, 63, 219, 228], "the": [9, 19, 37, 100, 126, 130, 181, 203, 277], "literature": [10], "for": [11, 257], "classification": [12, 28, 65, 199], "problems.": [13], "However,": [14], "integrating": [15], "these": [16, 144], "has": [18], "potential": [20], "to": [21, 58, 78], "enhance": [22], "their": [23], "combined": [24, 127], "power": [25], "improving": [27], "performance.": [29, 200], "We": [30], "present": [31], "a": [32, 40, 45, 49, 54, 244], "novel": [33], "approach": [34, 68, 192, 233], "that": [35, 74, 161], "combines": [36], "eigenanalysis": [38], "of": [39, 85, 94, 129, 205, 279], "covariance": [41, 253], "matrix": [42, 51], "evaluated": [43, 52], "on": [44, 53, 236], "training": [46], "set": [47], "with": [48, 221], "deep": [55, 304], "learning": [56], "model": [57], "achieve": [59, 138], "optimal": [60, 139, 263], "class": [61, 116, 140, 264], "separability": [62, 141, 227, 265], "binary": [64, 259], "tasks.": [66], "Our": [67, 167], "is": [69], "substantiated": [70], "by": [71, 171, 214, 266], "formal": [72], "proofs": [73], "establish": [75], "its": [76], "capability": [77], "maximize": [79], "between-class": [80], "mean": [81], "distance": [82], "(the": [83, 92], "concept": [84, 93], "separation": [86, 174, 272], ")": [87], "minimize": [89], "within-class": [90], "variances": [91], "compactness": [95, 176], "),": [96], "which": [97, 184, 224], "together": [98], "define": [99], "two": [101], "linear": [102, 226], "discriminant": [103], "analysis": [104], "(LDA)": [105], "criteria,": [106, 177, 208, 269], "particularly": [107], "under": [108], "ideal": [109, 280], "data": [110, 124, 249], "conditions": [111], "such": [112], "as": [113, 142], "isotropy": [114, 284], "around": [115], "means": [117], "dominant": [119, 286], "leading": [120, 287], "eigenvalues.": [121, 288], "By": [122], "projecting": [123], "into": [125, 303], "space": [128], "most": [131], "relevant": [132], "eigendirections": [133], "from": [134], "both": [135, 173, 206, 271], "matrices,": [136], "we": [137], "per": [143], "LDA": [145, 207, 212, 268], "criteria.": [146], "Empirical": [147, 290], "validation": [148, 291], "across": [149], "neural": [150, 305], "health": [152], "datasets": [153], "consistently": [154], "supports": [155], "our": [156, 162, 209, 232], "theoretical": [157], "framework": [158], "demonstrates": [160], "method": [163, 168, 210, 251], "outperforms": [164, 211], "established": [165], "methods.": [166, 299], "stands": [169], "out": [170], "addressing": [172], "unlike": [178], "PCA": [179], "method,": [183], "predominantly": [185], "emphasize": [186], "one": [187], "criterion": [188], "each.": [189], "This": [190], "comprehensive": [191], "captures": [193], "intricate": [194], "patterns": [195], "relationships,": [197], "enhancing": [198], "Furthermore,": [201], "through": [202, 309], "utilization": [204], "itself": [213], "leveraging": [215, 267], "higher-dimensional": [216], "feature": [217], "spaces,": [218], "accordance": [220], "Cover\u2019s": [222], "theorem,": [223], "favors": [225], "higher": [229], "dimensions.": [230], "Additionally,": [231], "sheds": [234], "light": [235], "complex": [237], "DNN": [238], "decision-making,": [239], "rendering": [240], "them": [241], "comprehensible": [242], "within": [243], "2D": [245], "space.": [246], "\u2022": [247, 261, 275, 289, 300], "Novel": [248], "projection": [250], "combining": [252], "enhanced": [258], "classification.": [260], "Achieves": [262], "emphasizing": [270], "compactness.": [274], "Highlights": [276], "importance": [278], "dataset": [281], "conditions,": [282], "including": [283], "shows": [292], "superior": [293], "performance": [294], "over": [295], "traditional": [296], "dimensionality": [297], "reduction": [298], "Offers": [301], "insights": [302], "network": [306], "decision-making": [307], "processes": [308], "eigenanalysis.": [310]}, "primary_topic": {"id": "https://openalex.org/T13141", "display_name": "Statistical Methods and Applications", "score": 0.7265999913215637, "subfield": {"id": "https://openalex.org/subfields/2613", "display_name": "Statistics and Probability"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T13141", "display_name": "Statistical Methods and Applications", "score": 0.7265999913215637, "subfield": {"id": "https://openalex.org/subfields/2613", "display_name": "Statistics and Probability"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10057", "display_name": "Face and Expression Recognition", "score": 0.12600000202655792, "subfield": {"id": "https://openalex.org/subfields/1707", "display_name": "Computer Vision and Pattern Recognition"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11396", "display_name": "Artificial Intelligence in Healthcare", "score": 0.033399999141693115, "subfield": {"id": "https://openalex.org/subfields/3605", "display_name": "Health Information Management"}, "field": {"id": "https://openalex.org/fields/36", "display_name": "Health Professions"}, "domain": {"id": "https://openalex.org/domains/4", "display_name": "Health Sciences"}}], "primary_location": {"id": "doi:10.1016/j.compbiomed.2025.109985", "is_oa": true, "landing_page_url": "https://doi.org/10.1016/j.compbiomed.2025.109985", "pdf_url": null, "source": {"id": "https://openalex.org/S44278595", "display_name": "Computers in Biology and Medicine", "issn_l": "0010-4825", "issn": ["0010-4825", "1879-0534"], "is_oa": false, "is_in_doaj": false, "is_core": true, "listed_in": ["cwts-core", "doyens", "jufo-2", "ki-jl-1", "medline", "norway-1"], "host_organization": "https://openalex.org/P4310320990", "host_organization_name": "Elsevier BV", "host_organization_lineage": ["https://openalex.org/P4310320990"], "host_organization_lineage_names": ["Elsevier BV"], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Computers in Biology and Medicine", "raw_type": "journal-article"}, "cited_by_count": 1, "is_retracted": false}, {"id": "https://openalex.org/W4412773805", "doi": "https://doi.org/10.36227/techrxiv.175393256.68603080/v1", "title": "Rule-based Tensor Mutations Embedded within LLMs for Low-Cost Mathematical Computation", "publication_year": 2025, "publication_date": "2025-07-31", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5119129405", "display_name": "Srikrishna Ayyalasomayajula", "orcid": "https://orcid.org/0009-0002-9144-5496", "observed_orcids": ["https://orcid.org/0009-0002-9144-5496"]}, "institutions": [], "countries": [], "is_corresponding": true, "raw_author_name": "Srikrishna Ayyalasomayajula", "raw_affiliation_strings": ["Plano , Texas"], "raw_orcid": "https://orcid.org/0009-0002-9144-5496", "affiliations": [{"raw_affiliation_string": "Plano , Texas", "institution_ids": []}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "have": [4], "demonstrated": [5], "remarkable": [6], "proficiency": [7], "in": [8], "natural": [9], "language": [10], "tasks": [11], "but": [12], "remain": [13], "inefficient": [14], "and": [15, 44, 81, 120, 126], "error-prone": [16], "when": [17], "performing": [18], "deterministic": [19, 108], "mathematical": [20, 26, 37, 96, 122], "computations.": [21], "Existing": [22], "approaches": [23], "to": [24], "improving": [25, 121], "reasoning": [27, 97], "rely": [28], "on": [29, 36, 124], "external": [30], "symbolic": [31], "engines": [32], "or": [33], "extensive": [34], "fine-tuning": [35], "corpora,": [38], "both": [39], "of": [40, 59, 87], "which": [41], "introduce": [42], "latency": [43, 119], "scalability": [45], "challenges.": [46], "This": [47], "paper": [48], "proposes": [49], "a": [50], "novel": [51], "architectural": [52], "enhancement": [53], "for": [54, 112], "transformer-based": [55], "LLMs:": [56], "the": [57, 66, 84, 88, 100], "embedding": [58, 85], "deterministic,": [60], "rule-based": [61], "tensor": [62, 74, 114], "mutations": [63], "directly": [64], "within": [65], "model's": [67], "internal": [68], "computational": [69], "graph.": [70], "By": [71], "implementing": [72], "fixed-index": [73], "operations-such": [75], "as": [76], "arithmetic": [77, 125], "functions,": [78], "binary": [79], "operations,": [80], "matrix": [82], "computations-within": [83], "space": [86], "Llama": [89], "3": [90], "3B": [91], "model,": [92], "we": [93], "enable": [94], "low-latency": [95], "without": [98], "modifying": [99], "core": [101], "probabilistic": [102], "architecture.": [103], "The": [104], "proposed": [105], "system": [106], "leverages": [107], "computation": [109], "pathways": [110], "optimized": [111], "GPU": [113], "cores,": [115], "significantly": [116], "reducing": [117], "inference": [118], "accuracy": [123], "linear": [127], "algebra": [128], "tasks.": [129]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.6333000063896179, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.6333000063896179, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.0786999985575676, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.062300000339746475, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.36227/techrxiv.175393256.68603080/v1", "is_oa": true, "landing_page_url": "https://doi.org/10.36227/techrxiv.175393256.68603080/v1", "pdf_url": "https://www.techrxiv.org/doi/pdf/10.36227/techrxiv.175393256.68603080", "source": null, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "acceptedVersion", "is_accepted": true, "is_published": false, "raw_source_name": null, "raw_type": "posted-content"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4409517555", "doi": "https://doi.org/10.1080/13658816.2025.2490701", "title": "Evaluating and enhancing spatial cognition abilities of large language models", "publication_year": 2025, "publication_date": "2025-04-15", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5046115734", "display_name": "Anran Yang", "orcid": "https://orcid.org/0000-0001-6106-759X", "observed_orcids": ["https://orcid.org/0000-0001-6106-759X"]}, "institutions": [{"id": "https://openalex.org/I170215575", "display_name": "National University of Defense Technology", "ror": "https://ror.org/05d2yfz11", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I170215575"]}, {"id": "https://openalex.org/I4210120125", "display_name": "Tianjin Institute of Advanced Technology", "ror": "https://ror.org/02tthqe19", "country_code": "CN", "type": "facility", "lineage": ["https://openalex.org/I4210120125"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Anran Yang", "raw_affiliation_strings": ["College of Electronic Science and Technology, National University of Defense Technology", "Tianjin Institute of Advanced Technology"], "raw_orcid": "https://orcid.org/0000-0001-6106-759X", "affiliations": [{"raw_affiliation_string": "College of Electronic Science and Technology, National University of Defense Technology", "institution_ids": ["https://openalex.org/I170215575"]}, {"raw_affiliation_string": "Tianjin Institute of Advanced Technology", "institution_ids": ["https://openalex.org/I4210120125"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5030330287", "display_name": "Cheng Fu", "orcid": "https://orcid.org/0000-0003-1562-7941", "observed_orcids": ["https://orcid.org/0000-0003-1562-7941"]}, "institutions": [{"id": "https://openalex.org/I202697423", "display_name": "University of Zurich", "ror": "https://ror.org/02crff812", "country_code": "CH", "type": "education", "lineage": ["https://openalex.org/I202697423"]}, {"id": "https://openalex.org/I4210151987", "display_name": "Ministry of Agriculture and Rural Affairs", "ror": "https://ror.org/05ckt8b96", "country_code": "CN", "type": "government", "lineage": ["https://openalex.org/I4210127390", "https://openalex.org/I4210151987"]}, {"id": "https://openalex.org/I52158045", "display_name": "China Agricultural University", "ror": "https://ror.org/04v3ywz14", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I52158045"]}], "countries": ["CH", "CN"], "is_corresponding": false, "raw_author_name": "Cheng Fu", "raw_affiliation_strings": ["College of Information and Electrical Engineering, China Agricultural University", "Department of Geography, University of Zurich", "Key Laboratory of Agricultural Machinery Monitoring and Big Data Applications, Ministry of Agriculture and Rural Affairs"], "raw_orcid": "https://orcid.org/0000-0003-1562-7941", "affiliations": [{"raw_affiliation_string": "College of Information and Electrical Engineering, China Agricultural University", "institution_ids": ["https://openalex.org/I52158045"]}, {"raw_affiliation_string": "Department of Geography, University of Zurich", "institution_ids": ["https://openalex.org/I202697423"]}, {"raw_affiliation_string": "Key Laboratory of Agricultural Machinery Monitoring and Big Data Applications, Ministry of Agriculture and Rural Affairs", "institution_ids": ["https://openalex.org/I4210151987"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5025161965", "display_name": "Qingren Jia", "orcid": "https://orcid.org/0000-0002-3741-5897", "observed_orcids": ["https://orcid.org/0000-0002-3741-5897"]}, "institutions": [{"id": "https://openalex.org/I170215575", "display_name": "National University of Defense Technology", "ror": "https://ror.org/05d2yfz11", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I170215575"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Qingren Jia", "raw_affiliation_strings": ["College of Electronic Science and Technology, National University of Defense Technology"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "College of Electronic Science and Technology, National University of Defense Technology", "institution_ids": ["https://openalex.org/I170215575"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5012133509", "display_name": "Weihua Dong", "orcid": "https://orcid.org/0000-0001-6097-7946", "observed_orcids": ["https://orcid.org/0000-0001-6097-7946"]}, "institutions": [{"id": "https://openalex.org/I25254941", "display_name": "Beijing Normal University", "ror": "https://ror.org/022k4wk35", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I25254941"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Weihua Dong", "raw_affiliation_strings": ["State Key Laboratory of Earth Surface Processes and Hazards Risk Governance, Faculty of Geographical Science, Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University"], "raw_orcid": "https://orcid.org/0000-0001-6097-7946", "affiliations": [{"raw_affiliation_string": "State Key Laboratory of Earth Surface Processes and Hazards Risk Governance, Faculty of Geographical Science, Advanced Interdisciplinary Institute of Satellite Applications, Beijing Normal University", "institution_ids": ["https://openalex.org/I25254941"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5101154782", "display_name": "Mengyu Ma", "orcid": "https://orcid.org/0009-0001-0897-7823", "observed_orcids": ["https://orcid.org/0009-0001-0897-7823"]}, "institutions": [{"id": "https://openalex.org/I170215575", "display_name": "National University of Defense Technology", "ror": "https://ror.org/05d2yfz11", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I170215575"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Mengyu Ma", "raw_affiliation_strings": ["College of Electronic Science and Technology, National University of Defense Technology"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "College of Electronic Science and Technology, National University of Defense Technology", "institution_ids": ["https://openalex.org/I170215575"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5100353589", "display_name": "Hao Chen", "orcid": "https://orcid.org/0000-0002-7880-3394", "observed_orcids": ["https://orcid.org/0000-0002-7880-3394", "https://orcid.org/0009-0002-6559-1043", "https://orcid.org/0009-0008-1632-7581", "https://orcid.org/0000-0003-1887-7890"]}, "institutions": [{"id": "https://openalex.org/I170215575", "display_name": "National University of Defense Technology", "ror": "https://ror.org/05d2yfz11", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I170215575"]}], "countries": ["CN"], "is_corresponding": true, "raw_author_name": "Hao Chen", "raw_affiliation_strings": ["College of Electronic Science and Technology, National University of Defense Technology"], "raw_orcid": "https://orcid.org/0000-0002-7880-3394", "affiliations": [{"raw_affiliation_string": "College of Electronic Science and Technology, National University of Defense Technology", "institution_ids": ["https://openalex.org/I170215575"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5069430511", "display_name": "Fei Yang", "orcid": "https://orcid.org/0000-0002-0717-7074", "observed_orcids": ["https://orcid.org/0000-0002-0717-7074"]}, "institutions": [{"id": "https://openalex.org/I170215575", "display_name": "National University of Defense Technology", "ror": "https://ror.org/05d2yfz11", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I170215575"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Fei Yang", "raw_affiliation_strings": ["College of Electronic Science and Technology, National University of Defense Technology"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "College of Electronic Science and Technology, National University of Defense Technology", "institution_ids": ["https://openalex.org/I170215575"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5101413395", "display_name": "Hui Wu", "orcid": "https://orcid.org/0000-0002-8881-3203", "observed_orcids": ["https://orcid.org/0000-0002-8881-3203", "https://orcid.org/0009-0005-8247-7260"]}, "institutions": [{"id": "https://openalex.org/I170215575", "display_name": "National University of Defense Technology", "ror": "https://ror.org/05d2yfz11", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I170215575"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Hui Wu", "raw_affiliation_strings": ["College of Electronic Science and Technology, National University of Defense Technology"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "College of Electronic Science and Technology, National University of Defense Technology", "institution_ids": ["https://openalex.org/I170215575"]}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "demonstrate": [4], "various": [5, 35], "capabilities": [6], "previously": [7], "considered": [8], "unique": [9], "to": [10, 17, 72, 107], "humans.": [11], "However,": [12], "current": [13], "evidence": [14], "is": [15], "insufficient": [16], "determine": [18], "whether": [19], "LLMs": [20, 77, 102, 149], "have": [21, 43], "developed": [22], "spatial": [23, 51, 66, 84, 112, 141, 153, 161], "cognition,": [24, 154], "a": [25, 62, 94, 122, 128], "fundamental": [26], "aspect": [27], "of": [28, 56, 83, 121, 180, 195], "human": [29], "cognition": [30, 52, 67], "underpinning": [31], "logical-mathematical": [32], "reasoning": [33], "and": [34, 79, 88, 168], "other": [36], "skills.": [37], "Previous": [38], "studies": [39], "on": [40, 46, 132], "this": [41], "topic": [42], "primarily": [44], "concentrated": [45], "small-scale": [47], "perceptions,": [48], "leaving": [49], "the": [50, 54, 119, 181, 185, 196], "within": [53], "context": [55], "GIScience": [57], "largely": [58], "unexamined.": [59], "We": [60], "introduce": [61], "benchmark": [63], "that": [64, 126, 147], "evaluates": [65], "abilities": [68], "across": [69, 159], "seven": [70], "categories": [71], "systematically": [73], "assess": [74], "how": [75], "well": [76], "process": [78], "generate": [80], "three": [81], "types": [82], "knowledge:": [85], "landmark,": [86], "route,": [87], "survey": [89, 169], "knowledge.": [90, 170], "Furthermore,": [91], "we": [92], "propose": [93], "tool-augmented": [95], "approach": [96, 188], "named": [97], "Hybrid": [98, 186], "Mind,": [99], "which": [100], "integrates": [101], "with": [103], "deterministic": [104], "GIS": [105], "algorithms": [106], "enhance": [108], "their": [109, 155], "performance": [110, 156], "in": [111, 139, 165], "cognitive": [113, 162], "tasks.": [114], "The": [115, 171], "core": [116], "idea": [117], "involves": [118], "implementation": [120], "mental": [123], "map": [124, 130], "builder": [125], "generates": [127], "quantitative": [129], "based": [131], "segmented": [133], "qualitative": [134], "constraints,": [135], "overcoming": [136], "LLMs\u2019": [137], "fallacies": [138], "synthesizing": [140], "information.": [142], "Our": [143], "experimental": [144], "results": [145], "revealed": [146], "although": [148], "exhibited": [150], "potential": [151], "for": [152], "was": [157], "poor": [158], "most": [160], "tasks,": [163], "particularly": [164], "constructing": [166], "route": [167], "leading": [172], "model,": [173], "GPT-4-turbo,": [174], "correctly": [175, 192], "answered": [176], "fewer": [177], "than": [178], "one-fourth": [179], "questions.": [182, 197], "In": [183], "contrast,": [184], "Mind": [187], "significantly": [189], "improved": [190], "performance,": [191], "solving": [193], "70.48%": [194]}, "primary_topic": {"id": "https://openalex.org/T11904", "display_name": "Spatial Cognition and Navigation", "score": 0.7871000170707703, "subfield": {"id": "https://openalex.org/subfields/2203", "display_name": "Automotive Engineering"}, "field": {"id": "https://openalex.org/fields/22", "display_name": "Engineering"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T11904", "display_name": "Spatial Cognition and Navigation", "score": 0.7871000170707703, "subfield": {"id": "https://openalex.org/subfields/2203", "display_name": "Automotive Engineering"}, "field": {"id": "https://openalex.org/fields/22", "display_name": "Engineering"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10757", "display_name": "Geographic Information Systems Studies", "score": 0.06679999828338623, "subfield": {"id": "https://openalex.org/subfields/3305", "display_name": "Geography, Planning and Development"}, "field": {"id": "https://openalex.org/fields/33", "display_name": "Social Sciences"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.02800000086426735, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "doi:10.1080/13658816.2025.2490701", "is_oa": true, "landing_page_url": "https://doi.org/10.1080/13658816.2025.2490701", "pdf_url": null, "source": {"id": "https://openalex.org/S4210181446", "display_name": "International Journal of Geographical Information Systems", "issn_l": "0269-3798", "issn": ["0269-3798", "1362-3087"], "is_oa": false, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/P4310320547", "host_organization_name": "Taylor & Francis", "host_organization_lineage": ["https://openalex.org/P4310320547", "https://openalex.org/P4310320449"], "host_organization_lineage_names": ["Taylor & Francis", "Informa"], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "International Journal of Geographical Information Science", "raw_type": "journal-article"}, "cited_by_count": 22, "is_retracted": false}, {"id": "https://openalex.org/W4417354674", "doi": "https://doi.org/10.48550/arxiv.2506.10674", "title": "TeleMath: A Benchmark for Large Language Models in Telecom Mathematical Problem Solving", "publication_year": 2025, "publication_date": "2025-06-12", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5120813430", "display_name": "Vincenzo Colle", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Colle, Vincenzo", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5028270135", "display_name": "Mohamed Sana", "orcid": "https://orcid.org/0000-0001-5644-5840", "observed_orcids": ["https://orcid.org/0000-0001-5644-5840"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Sana, Mohamed", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5048382270", "display_name": "Nicola Piovesan", "orcid": "https://orcid.org/0000-0001-5397-5248", "observed_orcids": ["https://orcid.org/0000-0001-5397-5248"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Piovesan, Nicola", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5045740906", "display_name": "Antonio De Domenico", "orcid": "https://orcid.org/0000-0003-1229-4045", "observed_orcids": ["https://orcid.org/0000-0003-1229-4045"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "De Domenico, Antonio", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5064737739", "display_name": "Fadhel Ayed", "orcid": "https://orcid.org/0009-0008-4234-7682", "observed_orcids": ["https://orcid.org/0009-0008-4234-7682"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Ayed, Fadhel", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5056145687", "display_name": "Merouane Abdelkader Debbah", "orcid": "https://orcid.org/0000-0001-8941-8080", "observed_orcids": ["https://orcid.org/0000-0001-8941-8080"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Debbah, Merouane", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"The": [0, 119], "increasing": [1], "adoption": [2], "of": [3, 14, 32, 93, 112, 121, 125, 156], "artificial": [4], "intelligence": [5], "in": [6, 11, 34, 72, 79, 95], "telecommunications": [7, 81, 97], "has": [8], "raised": [9], "interest": [10], "the": [12, 30, 62, 80, 96, 102, 166, 169], "capability": [13], "Large": [15], "Language": [16], "Models": [17], "(LLMs)": [18], "to": [19, 68, 172], "address": [20, 56], "domain-specific,": [21], "mathematically": [22], "intensive": [23], "tasks.": [24], "Although": [25], "recent": [26, 137], "advancements": [27], "have": [28, 164], "improved": [29], "performance": [31, 50, 71, 131], "LLMs": [33, 127], "general": [35], "mathematical": [36, 74, 142], "reasoning,": [37], "their": [38], "effectiveness": [39], "within": [40], "specialized": [41], "domains,": [42], "such": [43], "as": [44], "signal": [45], "processing,": [46], "network": [47], "optimization,": [48], "and": [49, 168, 176], "analysis,": [51], "remains": [52], "largely": [53], "unexplored.": [54], "To": [55], "this": [57], "gap,": [58], "we": [59], "introduce": [60], "TeleMath,": [61], "first": [63], "benchmark": [64], "dataset": [65, 167], "specifically": [66], "designed": [67, 140], "evaluate": [69], "LLM": [70], "solving": [73], "problems": [75, 113], "with": [76, 152, 160], "numerical": [77], "solutions": [78], "domain.": [82], "Comprising": [83], "500": [84], "question-answer": [85], "(QnA)": [86], "pairs,": [87], "TeleMath": [88, 133], "covers": [89], "a": [90, 109, 122, 153], "wide": [91, 123], "spectrum": [92], "topics": [94], "field.": [98], "This": [99], "paper": [100], "outlines": [101], "proposed": [103], "QnAs": [104], "generation": [105], "pipeline,": [106], "starting": [107], "from": [108], "selected": [110], "seed": [111], "crafted": [114], "by": [115, 136], "Subject": [116], "Matter": [117], "Experts.": [118], "evaluation": [120, 170], "range": [124], "open-source": [126], "reveals": [128], "that": [129], "best": [130], "on": [132], "is": [134], "achieved": [135], "models": [138], "explicitly": [139], "for": [141], "or": [143], "logical": [144], "reasoning.": [145], "In": [146], "contrast,": [147], "general-purpose": [148], "models,": [149], "even": [150], "those": [151], "large": [154], "number": [155], "parameters,": [157], "often": [158], "struggle": [159], "these": [161], "challenges.": [162], "We": [163], "released": [165], "code": [171], "ease": [173], "result": [174], "reproducibility": [175], "support": [177], "future": [178], "research.": [179]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8651999831199646, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.8651999831199646, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.04149999842047691, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13904", "display_name": "Artificial Intelligence Applications", "score": 0.021199999377131462, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2506.10674", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2506.10674", "pdf_url": "https://arxiv.org/pdf/2506.10674", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4416398373", "doi": "https://doi.org/10.48550/arxiv.2510.07748", "title": "Haibu Mathematical-Medical Intelligent Agent:Enhancing Large Language Model Reliability in Medical Tasks via Verifiable Reasoning Chains", "publication_year": 2025, "publication_date": "2025-10-09", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5027554813", "display_name": "Y. Zhang", "orcid": "https://orcid.org/0000-0002-0682-0961", "observed_orcids": ["https://orcid.org/0000-0002-0682-0961"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Zhang, Yilun", "raw_affiliation_strings": ["Zhejiang Qiushi Institute of Mathematical Medicine"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Zhejiang Qiushi Institute of Mathematical Medicine", "institution_ids": []}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5101478809", "display_name": "De-Xing Kong", "orcid": "https://orcid.org/0000-0003-3289-8954", "observed_orcids": ["https://orcid.org/0000-0003-3289-8954"]}, "institutions": [{"id": "https://openalex.org/I76130692", "display_name": "Zhejiang University", "ror": "https://ror.org/00a2xv884", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I76130692"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Kong, Dexing", "raw_affiliation_strings": ["School of Mathematical Sciences , Zhejiang University , Zhejiang Hangzhou , China"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "School of Mathematical Sciences , Zhejiang University , Zhejiang Hangzhou , China", "institution_ids": ["https://openalex.org/I76130692"]}]}], "abstract_inverted_index": {"Large": [0], "Language": [1], "Models": [2], "(LLMs)": [3], "show": [4], "promise": [5], "in": [6, 19, 194], "medicine": [7], "but": [8], "are": [9], "prone": [10], "to": [11, 72, 105, 156], "factual": [12], "and": [13, 68, 121, 183], "logical": [14, 66], "errors,": [15], "which": [16, 82], "is": [17, 61, 78, 154, 175], "unacceptable": [18], "this": [20], "high-stakes": [21], "field.": [22], "To": [23], "address": [24], "this,": [25], "we": [26], "introduce": [27], "the": [28, 150, 165], "\"Haibu": [29], "Mathematical-Medical": [30], "Intelligent": [31], "Agent\"": [32], "(MMIA),": [33], "an": [34, 132], "LLM-driven": [35], "architecture": [36], "that": [37], "ensures": [38], "reliability": [39], "through": [40], "a": [41, 106, 139, 176], "formally": [42], "verifiable": [43, 172], "reasoning": [44, 59, 85, 104, 173], "process.": [45], "MMIA": [46, 112, 130], "recursively": [47], "breaks": [48], "down": [49], "complex": [50], "medical": [51, 122], "tasks": [52, 90], "into": [53], "atomic,": [54], "evidence-based": [55], "steps.": [56], "This": [57], "entire": [58], "chain": [60], "then": [62, 92], "automatically": [63], "audited": [64], "for": [65, 191], "coherence": [67], "evidence": [69], "traceability,": [70], "similar": [71], "theorem": [73], "proving.": [74], "A": [75], "key": [76], "innovation": [77], "MMIA's": [79, 171], "\"bootstrapping\"": [80], "mode,": [81], "stores": [83], "validated": [84, 111], "chains": [86], "as": [87, 164], "\"theorems.\"": [88], "Subsequent": [89], "can": [91], "be": [93], "efficiently": [94], "solved": [95], "using": [96, 125], "Retrieval-Augmented": [97], "Generation": [98], "(RAG),": [99], "shifting": [100], "from": [101], "costly": [102], "first-principles": [103], "low-cost": [107], "verification": [108], "model.": [109], "We": [110], "across": [113], "four": [114], "healthcare": [115], "administration": [116], "domains,": [117], "including": [118], "DRG/DIP": [119], "audits": [120], "insurance": [123], "adjudication,": [124], "expert-validated": [126], "benchmarks.": [127], "Results": [128], "showed": [129], "achieved": [131], "error": [133], "detection": [134], "rate": [135, 142], "exceeding": [136], "98%": [137], "with": [138], "false": [140], "positive": [141], "below": [143], "1%,": [144], "significantly": [145], "outperforming": [146], "baseline": [147], "LLMs.": [148], "Furthermore,": [149], "RAG": [151], "matching": [152], "mode": [153], "projected": [155], "reduce": [157], "average": [158], "processing": [159], "costs": [160], "by": [161], "approximately": [162], "85%": [163], "knowledge": [166], "base": [167], "matures.": [168], "In": [169], "conclusion,": [170], "framework": [174], "significant": [177], "step": [178], "toward": [179], "creating": [180], "trustworthy,": [181], "transparent,": [182], "cost-effective": [184], "AI": [185], "systems,": [186], "making": [187], "LLM": [188], "technology": [189], "viable": [190], "critical": [192], "applications": [193], "medicine.": [195]}, "primary_topic": {"id": "https://openalex.org/T11636", "display_name": "Artificial Intelligence in Healthcare and Education", "score": 0.9448000192642212, "subfield": {"id": "https://openalex.org/subfields/2718", "display_name": "Health Informatics"}, "field": {"id": "https://openalex.org/fields/27", "display_name": "Medicine"}, "domain": {"id": "https://openalex.org/domains/4", "display_name": "Health Sciences"}}, "topics": [{"id": "https://openalex.org/T11636", "display_name": "Artificial Intelligence in Healthcare and Education", "score": 0.9448000192642212, "subfield": {"id": "https://openalex.org/subfields/2718", "display_name": "Health Informatics"}, "field": {"id": "https://openalex.org/fields/27", "display_name": "Medicine"}, "domain": {"id": "https://openalex.org/domains/4", "display_name": "Health Sciences"}}, {"id": "https://openalex.org/T12026", "display_name": "Explainable Artificial Intelligence (XAI)", "score": 0.028999999165534973, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13295", "display_name": "Safety Systems Engineering in Autonomy", "score": 0.003700000001117587, "subfield": {"id": "https://openalex.org/subfields/2213", "display_name": "Safety, Risk, Reliability and Quality"}, "field": {"id": "https://openalex.org/fields/22", "display_name": "Engineering"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2510.07748", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2510.07748", "pdf_url": "https://arxiv.org/pdf/2510.07748", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by-nc-nd", "license_id": "https://openalex.org/licenses/cc-by-nc-nd", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4406840443", "doi": "https://doi.org/10.48550/arxiv.2501.14002", "title": "Advancing Mathematical Reasoning in Language Models: The Impact of Problem-Solving Data, Data Synthesis Methods, and Training Stages", "publication_year": 2025, "publication_date": "2025-01-23", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5101718220", "display_name": "Zui Chen", "orcid": "https://orcid.org/0009-0008-3488-7312", "observed_orcids": ["https://orcid.org/0009-0008-3488-7312"]}, "institutions": [{"id": "https://openalex.org/I159948400", "display_name": "Jinan University", "ror": "https://ror.org/02xe5ns62", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I159948400"]}, {"id": "https://openalex.org/I30809798", "display_name": "ShanghaiTech University", "ror": "https://ror.org/030bhh786", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I30809798"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Chen, Zui", "raw_affiliation_strings": ["Jinan University", "ShanghaiTech University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Jinan University", "institution_ids": ["https://openalex.org/I159948400"]}, {"raw_affiliation_string": "ShanghaiTech University", "institution_ids": ["https://openalex.org/I30809798"]}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5078859530", "display_name": "Tianqiao Liu", "orcid": "https://orcid.org/0000-0001-7997-3232", "observed_orcids": ["https://orcid.org/0000-0001-7997-3232", "https://orcid.org/0000-0002-6287-207X"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Liu, Tianqiao", "raw_affiliation_strings": ["TAL Education Group"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TAL Education Group", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": null, "display_name": "Tian, Mi", "orcid": null, "observed_orcids": []}, "institutions": [{"id": "https://openalex.org/I159948400", "display_name": "Jinan University", "ror": "https://ror.org/02xe5ns62", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I159948400"]}, {"id": "https://openalex.org/I32246829", "display_name": "Guangdong University of Education", "ror": "https://ror.org/0574der91", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I32246829"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Tian, Mi", "raw_affiliation_strings": ["Guangdong Institute of Smart Education , Jinan University , Guangzhou , China", "TAL Education Group"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Guangdong Institute of Smart Education , Jinan University , Guangzhou , China", "institution_ids": ["https://openalex.org/I159948400", "https://openalex.org/I32246829"]}, {"raw_affiliation_string": "TAL Education Group", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102738062", "display_name": "Qing Tong", "orcid": "https://orcid.org/0000-0002-8929-3043", "observed_orcids": ["https://orcid.org/0000-0002-8929-3043"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Tong, Qing", "raw_affiliation_strings": ["TAL Education Group"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "TAL Education Group", "institution_ids": []}]}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5102712272", "display_name": "Weiqi Luo", "orcid": "https://orcid.org/0000-0002-4153-9040", "observed_orcids": ["https://orcid.org/0000-0002-4153-9040"]}, "institutions": [{"id": "https://openalex.org/I159948400", "display_name": "Jinan University", "ror": "https://ror.org/02xe5ns62", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I159948400"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Luo, Weiqi", "raw_affiliation_strings": ["Jinan University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Jinan University", "institution_ids": ["https://openalex.org/I159948400"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5014957271", "display_name": "Zitao Liu", "orcid": "https://orcid.org/0000-0003-0491-307X", "observed_orcids": ["https://orcid.org/0000-0003-0491-307X"]}, "institutions": [{"id": "https://openalex.org/I159948400", "display_name": "Jinan University", "ror": "https://ror.org/02xe5ns62", "country_code": "CN", "type": "education", "lineage": ["https://openalex.org/I159948400"]}], "countries": ["CN"], "is_corresponding": false, "raw_author_name": "Liu, Zitao", "raw_affiliation_strings": ["Jinan University"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "Jinan University", "institution_ids": ["https://openalex.org/I159948400"]}]}], "abstract_inverted_index": {"Mathematical": [0], "reasoning": [1, 53, 107, 233], "remains": [2], "a": [3, 29, 242], "challenging": [4, 221], "area": [5], "for": [6, 42, 219, 229], "large": [7], "language": [8], "models": [9, 26], "(LLMs),": [10], "prompting": [11], "the": [12, 49, 79, 84, 104, 122, 137, 141, 147, 167, 185, 191, 206, 231], "development": [13, 240], "of": [14, 86, 235, 241], "math-specific": [15], "LLMs": [16], "such": [17], "as": [18], "LLEMMA,": [19], "DeepSeekMath,": [20], "and": [21, 37, 127, 149, 152], "Qwen2-Math,": [22], "among": [23], "others.": [24], "These": [25, 224], "typically": [27], "follow": [28], "two-stage": [30], "training": [31], "paradigm:": [32], "pre-training": [33, 57, 80], "with": [34, 39, 205], "math-related": [35], "corpora": [36, 114], "post-training": [38], "problem": [40], "datasets": [41], "supervised": [43], "fine-tuning": [44], "(SFT).": [45], "Despite": [46], "these": [47, 157], "efforts,": [48], "improvements": [50], "in": [51, 238], "mathematical": [52, 91, 106, 113, 169, 174, 232, 244], "achieved": [54], "through": [55], "continued": [56], "(CPT)": [58], "are": [59, 131], "often": [60], "less": [61], "significant": [62], "compared": [63, 171, 202], "to": [64, 156, 172, 203, 214], "those": [65], "obtained": [66], "via": [67], "SFT.": [68], "This": [69], "study": [70], "addresses": [71], "this": [72], "discrepancy": [73], "by": [74], "exploring": [75], "alternative": [76], "strategies": [77], "during": [78, 115], "phase,": [81], "focusing": [82], "on": [83], "use": [85], "problem-solving": [87, 101, 143, 163, 222], "data": [88, 102, 120, 144, 164, 180], "over": [89], "general": [90, 112, 173], "corpora.": [92, 175], "We": [93, 176], "investigate": [94], "three": [95], "primary": [96], "research": [97], "questions:": [98], "(1)": [99], "Can": [100], "enhance": [103], "model's": [105, 168], "capabilities": [108, 138, 170, 234], "more": [109, 220], "effectively": [110], "than": [111], "CPT?": [116], "(2)": [117], "Are": [118], "synthetic": [119], "from": [121, 140], "same": [123, 142, 207], "source": [124], "equally": [125], "effective,": [126], "which": [128, 209], "synthesis": [129, 181, 188], "methods": [130], "most": [132], "efficient?": [133], "(3)": [134], "How": [135], "do": [136], "developed": [139], "differ": [145], "between": [146], "CPT": [148, 204], "SFT": [150, 196], "stages,": [151], "what": [153], "factors": [154], "contribute": [155], "differences?": [158], "Our": [159], "findings": [160], "indicate": [161], "that": [162, 184], "significantly": [165], "enhances": [166], "also": [177], "identify": [178], "effective": [179], "methods,": [182], "demonstrating": [183], "tutorship": [186], "amplification": [187], "method": [189], "achieves": [190], "best": [192], "performance.": [193], "Furthermore,": [194], "while": [195], "facilitates": [197], "instruction-following": [198], "abilities,": [199], "it": [200], "underperforms": [201], "data,": [208], "can": [210], "be": [211], "partially": [212], "attributed": [213], "its": [215], "poor": [216], "learning": [217], "capacity": [218], "data.": [223], "insights": [225], "provide": [226], "valuable": [227], "guidance": [228], "optimizing": [230], "LLMs,": [236], "culminating": [237], "our": [239], "powerful": [243], "base": [245], "model": [246], "called": [247], "MathGPT-8B.": [248]}, "primary_topic": {"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7437000274658203, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10028", "display_name": "Topic Modeling", "score": 0.7437000274658203, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.06300000101327896, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10181", "display_name": "Natural Language Processing Techniques", "score": 0.02810000069439411, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2501.14002", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2501.14002", "pdf_url": "https://arxiv.org/pdf/2501.14002", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4417426882", "doi": "https://doi.org/10.48550/arxiv.2507.14330", "title": "Leveraging LLMs for Formal Software Requirements -- Challenges and Prospects", "publication_year": 2025, "publication_date": "2025-07-18", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5066209708", "display_name": "Arshad Beg", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Beg, Arshad", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": null, "display_name": "O'Donoghue, Diarmuid", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "O'Donoghue, Diarmuid", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5003866910", "display_name": "Rosemary Monahan", "orcid": "https://orcid.org/0000-0003-3886-4675", "observed_orcids": ["https://orcid.org/0000-0003-3886-4675"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Monahan, Rosemary", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Software": [0], "correctness": [1, 34], "is": [2, 45], "ensured": [3], "mathematically": [4], "through": [5], "formal": [6, 14], "verification,": [7], "which": [8], "involves": [9], "the": [10, 37, 40, 49, 72, 125], "resources": [11], "of": [12, 51, 113, 127], "generating": [13], "requirement": [15], "specifications": [16, 63, 129], "and": [17, 29, 58, 66, 82, 101, 120], "having": [18], "an": [19], "implementation": [20, 38], "that": [21], "must": [22], "be": [23], "verified.": [24], "Tools": [25], "such": [26], "as": [27], "model-checkers": [28], "theorem": [30], "provers": [31], "ensure": [32], "software": [33], "by": [35], "verifying": [36], "against": [39], "specification.": [41], "Formal": [42], "methods": [43], "deployment": [44], "regularly": [46], "enforced": [47], "in": [48, 124], "development": [50], "safety-critical": [52], "systems": [53], "e.g.": [54], "aerospace,": [55], "medical": [56], "devices": [57], "autonomous": [59], "systems.": [60], "Generating": [61], "these": [62], "from": [64, 91, 130], "informal": [65, 131], "ambiguous": [67], "natural": [68], "language": [69, 103], "requirements": [70], "remains": [71], "key": [73], "challenge.": [74], "Our": [75], "project,": [76], "VERIFAI^{1},": [77], "aims": [78], "to": [79, 85, 116], "investigate": [80], "automated": [81], "semi-automated": [83], "approaches": [84], "bridge": [86], "this": [87], "gap,": [88], "using": [89], "techniques": [90], "Natural": [92], "Language": [93], "Processing": [94], "(NLP),": [95], "ontology-based": [96], "domain": [97], "modelling,": [98], "artefact": [99], "reuse,": [100], "large": [102], "models": [104], "(LLMs).": [105], "This": [106], "position": [107], "paper": [108], "presents": [109], "a": [110], "preliminary": [111], "synthesis": [112], "relevant": [114], "literature": [115], "identify": [117], "recurring": [118], "challenges": [119], "prospective": [121], "research": [122], "directions": [123], "generation": [126], "verifiable": [128], "requirements.": [132]}, "primary_topic": {"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.7939000129699707, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.7939000129699707, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11450", "display_name": "Model-Driven Software Engineering Techniques", "score": 0.06509999930858612, "subfield": {"id": "https://openalex.org/subfields/1712", "display_name": "Software"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.0430000014603138, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2507.14330", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2507.14330", "pdf_url": "https://arxiv.org/pdf/2507.14330", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 1, "is_retracted": false}, {"id": "https://openalex.org/W4417426748", "doi": null, "title": "ProofCompass: Enhancing Specialized Provers with LLM Guidance", "publication_year": 2025, "publication_date": "2025-07-18", "type": "preprint", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5120832687", "display_name": "Nicolas Wischermann", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Wischermann, Nicolas", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5104453408", "display_name": "Claudio Mayrink Verdun", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Verdun, Claudio Mayrink", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "middle", "author": {"id": "https://openalex.org/A5043645877", "display_name": "Gabriel Poesia", "orcid": "https://orcid.org/0009-0003-7058-8428", "observed_orcids": ["https://orcid.org/0009-0003-7058-8428"]}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Poesia, Gabriel", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}, {"author_position": "last", "author": {"id": "https://openalex.org/A5120832690", "display_name": "Francesco Noseda", "orcid": null, "observed_orcids": []}, "institutions": [], "countries": [], "is_corresponding": false, "raw_author_name": "Noseda, Francesco", "raw_affiliation_strings": [], "raw_orcid": null, "affiliations": []}], "abstract_inverted_index": {"Language": [0, 67], "models": [1, 21, 34], "have": [2], "become": [3], "increasingly": [4], "powerful": [5], "tools": [6], "for": [7, 123], "formal": [8, 131], "mathematical": [9], "reasoning.": [10], "However,": [11], "most": [12], "existing": [13, 56], "approaches": [14], "rely": [15], "exclusively": [16], "on": [17], "either": [18], "large": [19, 33], "general-purpose": [20], "or": [22], "smaller": [23], "specialized": [24, 32, 57], "models,": [25], "each": [26], "with": [27, 64], "distinct": [28], "limitations,": [29], "while": [30, 109], "training": [31], "still": [35], "requires": [36], "significant": [37], "computational": [38, 51, 126], "resources.": [39], "This": [40], "paper": [41], "introduces": [42], "ProofCompass,": [43], "a": [44, 65], "novel": [45], "hybrid": [46], "methodology": [47], "that": [48], "achieves": [49], "remarkable": [50], "efficiency": [52, 127], "by": [53], "strategically": [54], "guiding": [55], "prover": [58], "methods,": [59], "such": [60], "as": [61], "DeepSeek-Prover-v1.5-RL": [62], "(DSP-v1.5)": [63], "Large": [66], "Model": [68], "(LLM)": [69], "without": [70], "requiring": [71], "additional": [72], "model": [73], "training.": [74], "The": [75], "LLM": [76], "provides": [77], "natural": [78], "language": [79], "proof": [80], "strategies": [81], "and": [82, 128], "analyzes": [83], "failed": [84], "attempts": [85, 113], "to": [86], "select": [87], "intermediate": [88], "lemmas,": [89], "enabling": [90], "effective": [91], "problem": [92], "decomposition.": [93], "On": [94], "the": [95, 121], "miniF2F": [96], "benchmark,": [97], "ProofCompass": [98], "demonstrates": [99], "substantial": [100], "resource": [101], "efficiency:": [102], "it": [103], "outperforms": [104], "DSP-v1.5": [105], "($54.9\\%": [106], "\\rightarrow": [107, 115], "55.3\\%$)": [108], "using": [110], "25x": [111], "fewer": [112], "($3200": [114], "128$).": [116], "Our": [117], "synergistic": [118], "approach": [119], "paves": [120], "way": [122], "simultaneously": [124], "improving": [125], "accuracy": [129], "in": [130], "theorem": [132], "proving.": [133]}, "primary_topic": {"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.6582000255584717, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T10126", "display_name": "Logic, programming, and type systems", "score": 0.6582000255584717, "subfield": {"id": "https://openalex.org/subfields/1702", "display_name": "Artificial Intelligence"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T13523", "display_name": "Mathematics, Computing, and Information Processing", "score": 0.17679999768733978, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T10142", "display_name": "Formal Methods in Verification", "score": 0.035100001841783524, "subfield": {"id": "https://openalex.org/subfields/1703", "display_name": "Computational Theory and Mathematics"}, "field": {"id": "https://openalex.org/fields/17", "display_name": "Computer Science"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}], "primary_location": {"id": "pmh:oai:arXiv.org:2507.14335", "is_oa": true, "landing_page_url": "http://arxiv.org/abs/2507.14335", "pdf_url": "https://arxiv.org/pdf/2507.14335", "source": {"id": "https://openalex.org/S4306400194", "display_name": "arXiv (Cornell University)", "issn_l": "2331-8422", "issn": ["2331-8422"], "is_oa": true, "is_in_doaj": false, "is_core": false, "listed_in": [], "host_organization": "https://openalex.org/I205783295", "host_organization_name": "Cornell University", "host_organization_lineage": ["https://openalex.org/I205783295"], "host_organization_lineage_names": [], "type": "repository"}, "license": null, "license_id": null, "version": "submittedVersion", "is_accepted": false, "is_published": false, "raw_source_name": null, "raw_type": "text"}, "cited_by_count": 0, "is_retracted": false}, {"id": "https://openalex.org/W4413958444", "doi": "https://doi.org/10.1017/psa.2025.10160", "title": "A Priori Knowledge in an Era of Computational Opacity: The Role of Artificial Intelligence in Mathematical Discovery", "publication_year": 2025, "publication_date": "2025-09-03", "type": "article", "language": "en", "authorships": [{"author_position": "first", "author": {"id": "https://openalex.org/A5028411525", "display_name": "Eamon Duede", "orcid": "https://orcid.org/0000-0002-3592-0478", "observed_orcids": ["https://orcid.org/0000-0002-3592-0478"]}, "institutions": [{"id": "https://openalex.org/I1282105669", "display_name": "Argonne National Laboratory", "ror": "https://ror.org/05gvnxz63", "country_code": "US", "type": "facility", "lineage": ["https://openalex.org/I1282105669", "https://openalex.org/I1330989302", "https://openalex.org/I39565521", "https://openalex.org/I40347166"]}, {"id": "https://openalex.org/I219193219", "display_name": "Purdue University West Lafayette", "ror": "https://ror.org/02dqehb95", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I219193219"]}], "countries": ["US"], "is_corresponding": true, "raw_author_name": "Eamon Duede", "raw_affiliation_strings": ["Argonne National Laboratory", "Purdue University"], "raw_orcid": "https://orcid.org/0000-0002-3592-0478", "affiliations": [{"raw_affiliation_string": "Argonne National Laboratory", "institution_ids": ["https://openalex.org/I1282105669"]}, {"raw_affiliation_string": "Purdue University", "institution_ids": ["https://openalex.org/I219193219"]}]}, {"author_position": "last", "author": {"id": "https://openalex.org/A5029112871", "display_name": "Kevin Davey", "orcid": "https://orcid.org/0000-0002-0791-7167", "observed_orcids": ["https://orcid.org/0000-0002-0791-7167"]}, "institutions": [{"id": "https://openalex.org/I40347166", "display_name": "University of Chicago", "ror": "https://ror.org/024mw5h28", "country_code": "US", "type": "education", "lineage": ["https://openalex.org/I40347166"]}], "countries": ["US"], "is_corresponding": true, "raw_author_name": "Kevin Davey", "raw_affiliation_strings": ["University of Chicago"], "raw_orcid": null, "affiliations": [{"raw_affiliation_string": "University of Chicago", "institution_ids": ["https://openalex.org/I40347166"]}]}], "abstract_inverted_index": {"Abstract": [0], "Can": [1], "we": [2, 15, 80], "acquire": [3], "a": [4, 21, 59, 68, 84], "priori": [5, 22, 60, 85], "mathematical": [6, 40, 61, 86], "knowledge": [7, 23, 62, 87], "from": [8, 28, 88], "the": [9, 25, 42, 92, 101], "outputs": [10], "of": [11, 24, 39, 44, 73], "computer": [12, 30], "programs?": [13], "Although": [14], "claim": [16], "Appel": [17], "and": [18, 50, 100], "Haken": [19], "acquired": [20], "four-color": [26], "theorem": [27], "their": [29], "program": [31], "insofar": [32], "as": [33], "it": [34], "merely": [35], "automated": [36], "human": [37, 71], "forms": [38, 72], "reasoning,": [41], "opacity": [43], "modern": [45], "large": [46], "language": [47], "models": [48], "(LLMs)": [49], "deep": [51], "neural": [52], "networks": [53], "(DNNs)": [54], "creates": [55], "obstacles": [56], "in": [57, 63], "obtaining": [58], "analogous": [64], "ways.": [65], "If,": [66], "however,": [67], "proof-checker": [69], "automating": [70], "proof-checking": [74], "is": [75], "attached": [76], "to": [77, 98], "such": [78], "machines,": [79], "can": [81], "indeed": [82], "obtain": [83], "them,": [89], "even": [90], "though": [91], "original": [93], "machines": [94], "are": [95], "entirely": [96], "opaque": [97], "us": [99], "outputted": [102], "proofs": [103], "cannot": [104], "be": [105], "surveyed": [106], "by": [107], "humans.": [108]}, "primary_topic": {"id": "https://openalex.org/T12170", "display_name": "History and Theory of Mathematics", "score": 0.4154999852180481, "subfield": {"id": "https://openalex.org/subfields/2614", "display_name": "Theoretical Computer Science"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, "topics": [{"id": "https://openalex.org/T12170", "display_name": "History and Theory of Mathematics", "score": 0.4154999852180481, "subfield": {"id": "https://openalex.org/subfields/2614", "display_name": "Theoretical Computer Science"}, "field": {"id": "https://openalex.org/fields/26", "display_name": "Mathematics"}, "domain": {"id": "https://openalex.org/domains/3", "display_name": "Physical Sciences"}}, {"id": "https://openalex.org/T11477", "display_name": "Epistemology, Ethics, and Metaphysics", "score": 0.28200000524520874, "subfield": {"id": "https://openalex.org/subfields/1211", "display_name": "Philosophy"}, "field": {"id": "https://openalex.org/fields/12", "display_name": "Arts and Humanities"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}, {"id": "https://openalex.org/T10258", "display_name": "Philosophy and Theoretical Science", "score": 0.0697999969124794, "subfield": {"id": "https://openalex.org/subfields/3205", "display_name": "Experimental and Cognitive Psychology"}, "field": {"id": "https://openalex.org/fields/32", "display_name": "Psychology"}, "domain": {"id": "https://openalex.org/domains/2", "display_name": "Social Sciences"}}], "primary_location": {"id": "doi:10.1017/psa.2025.10160", "is_oa": true, "landing_page_url": "https://doi.org/10.1017/psa.2025.10160", "pdf_url": "https://www.cambridge.org/core/services/aop-cambridge-core/content/view/0192BDB2814A219D9435A912786FE4CA/S0031824825101608a.pdf/div-class-title-philosophy-of-science-association-latex-guidelines-for-authors-div.pdf", "source": {"id": "https://openalex.org/S147691735", "display_name": "Philosophy of Science", "issn_l": "0031-8248", "issn": ["0031-8248", "1539-767X", "1540-4684"], "is_oa": false, "is_in_doaj": false, "is_core": true, "listed_in": ["cwts-core", "doyens", "erih-plus", "jufo-3", "norway-2"], "host_organization": "https://openalex.org/P4310311721", "host_organization_name": "Cambridge University Press", "host_organization_lineage": ["https://openalex.org/P4310311721", "https://openalex.org/P4310311702"], "host_organization_lineage_names": ["Cambridge University Press", "University of Cambridge"], "type": "journal"}, "license": "cc-by", "license_id": "https://openalex.org/licenses/cc-by", "version": "publishedVersion", "is_accepted": true, "is_published": true, "raw_source_name": "Philosophy of Science", "raw_type": "journal-article"}, "cited_by_count": 0, "is_retracted": false}], "group_by": []}