{"id":"https://openalex.org/W7166863113","doi":"https://doi.org/10.18653/v1/2026.findings-acl.936","title":"RAVR: Reference-Answer-guided Variational Reasoning for Large Language Models","display_name":"RAVR: Reference-Answer-guided Variational Reasoning for Large Language Models","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166863113","doi":"https://doi.org/10.18653/v1/2026.findings-acl.936"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.936","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.936","pdf_url":"https://aclanthology.org/2026.findings-acl.936.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.936.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5085380143","display_name":"Tianqianjin Lin","orcid":"https://orcid.org/0000-0002-4272-394X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tianqianjin Lin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139730104","display_name":"Xi Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xi Zhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139721311","display_name":"Xingyao Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xingyao Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023108787","display_name":"Rujiao Long","orcid":"https://orcid.org/0000-0003-1330-3193"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rujiao Long","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139769950","display_name":"Yi Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yi Xu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139816487","display_name":"Zhuoren Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhuoren Jiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139762870","display_name":"Wenbo Su","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wenbo Su","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139732377","display_name":"Bo Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bo Zheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.85187254,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"18757","last_page":"18775"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.6128000020980835,"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":0.6128000020980835,"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":0.16130000352859497,"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":0.053599998354911804,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/automated-reasoning","display_name":"Automated reasoning","score":0.4009999930858612},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.313400000333786},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.3122999966144562},{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.27959999442100525},{"id":"https://openalex.org/keywords/non-monotonic-logic","display_name":"Non-monotonic logic","score":0.2759000062942505}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5425999760627747},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4733000099658966},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.4009999930858612},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3452000021934509},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.313400000333786},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.3122999966144562},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.29010000824928284},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.27959999442100525},{"id":"https://openalex.org/C159032336","wikidata":"https://www.wikidata.org/wiki/Q2488768","display_name":"Non-monotonic logic","level":2,"score":0.2759000062942505},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.275299996137619},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.25679999589920044},{"id":"https://openalex.org/C179603123","wikidata":"https://www.wikidata.org/wiki/Q1941921","display_name":"Modeling language","level":3,"score":0.25270000100135803}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.936","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.936","pdf_url":"https://aclanthology.org/2026.findings-acl.936.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.936","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.936","pdf_url":"https://aclanthology.org/2026.findings-acl.936.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.5089432597160339}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166863113.pdf","grobid_xml":"https://content.openalex.org/works/W7166863113.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Reinforcement":[0],"learning":[1,46],"(RL)":[2],"can":[3,22,40,105],"refine":[4],"the":[5,20,33,56,65,75,80,93,122,126,171,178],"reasoning":[6,26,38,94,112,131,156,179],"abilities":[7],"of":[8,84,129,173,177],"large":[9],"language":[10],"models":[11,169],"(LLMs),":[12],"but":[13,50],"critically":[14],"depends":[15],"on":[16,121,140],"a":[17,97,158],"key":[18],"prerequisite:":[19],"LLM":[21],"already":[23],"generate":[24],"high-utility":[25],"paths":[27,39],"with":[28],"nonnegligible":[29],"probability.For":[30],"tasks":[31],"beyond":[32],"LLM's":[34],"current":[35],"competence,":[36],"such":[37],"be":[41],"hard":[42],"to":[43,99,109],"sample,":[44],"and":[45,117,167,175,189],"risks":[47],"reinforcing":[48],"familiar":[49],"suboptimal":[51],"reasoning.We":[52],"are":[53],"motivated":[54],"by":[55],"insight":[57],"from":[58],"cognitive":[59,82],"science":[60],"that":[61,95,103,119,153,182],"Why":[62],"is":[63,67,74],"this":[64,115,141],"answer":[66,123],"often":[68],"an":[69,150],"easier":[70],"question":[71,98],"than":[72],"What":[73],"answer,":[76],"as":[77,157],"it":[78],"avoids":[79],"heavy":[81],"load":[83],"openended":[85],"exploration,":[86],"opting":[87],"instead":[88],"for":[89,161],"explanatory":[90],"reconstruction-systematically":[91],"retracing":[92],"links":[96],"its":[100],"answer.We":[101],"show":[102],"LLMs":[104],"similarly":[106],"leverage":[107],"answers":[108],"derive":[110],"high-quality":[111],"paths.We":[113],"formalize":[114],"phenomenon":[116],"prove":[118],"conditioning":[120],"provably":[124],"increases":[125],"expected":[127],"utility":[128],"sampled":[130],"paths,":[132],"thereby":[133],"transforming":[134],"intractable":[135],"problems":[136],"into":[137],"learnable":[138],"ones.Building":[139],"insight,":[142],"we":[143],"introduce":[144],"RAVR":[145,183],"1":[146],"(Reference-Answer-guided":[147],"Variational":[148],"Reasoning),":[149],"end-to-end":[151],"framework":[152],"uses":[154],"answer-conditioned":[155],"variational":[159],"surrogate":[160],"question-only":[162],"reasoning.Experiments":[163],"across":[164],"12":[165],"benchmarks":[166],"4":[168],"demonstrate":[170],"effectiveness":[172],"RAVR,":[174],"analysis":[176],"behavior":[180],"shows":[181],"reduces":[184],"hesitation,":[185],"strengthens":[186],"conclusion":[187],"consolidation,":[188],"promotes":[190],"problemspecific":[191],"strategies":[192],"in":[193],"reasoning.":[194]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}