{"id":"https://openalex.org/W7147493658","doi":"https://doi.org/10.48550/arxiv.2603.28135","title":"CoT2-Meta: Budgeted Metacognitive Control for Test-Time Reasoning","display_name":"CoT2-Meta: Budgeted Metacognitive Control for Test-Time Reasoning","publication_year":2026,"publication_date":"2026-03-30","ids":{"openalex":"https://openalex.org/W7147493658","doi":"https://doi.org/10.48550/arxiv.2603.28135"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.28135","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.28135","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.28135","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5132667764","display_name":"Siyuan Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Siyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132634292","display_name":"Bo Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Bo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101301177","display_name":"Zikai Xiao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiao, Zikai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132587853","display_name":"Hailong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Hailong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132611835","display_name":"Xinlei Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Xinlei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132726309","display_name":"Rui Qian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qian, Rui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132628139","display_name":"Jiayu Qian","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qian, Jiayu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132655670","display_name":"Luqi Gong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gong, Luqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5132698519","display_name":"Yang Liu","orcid":"https://orcid.org/0000-0001-8552-0096"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Yang","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":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.41940000653266907,"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":0.41940000653266907,"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":0.12030000239610672,"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":0.06019999831914902,"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/inference","display_name":"Inference","score":0.6632000207901001},{"id":"https://openalex.org/keywords/metacognition","display_name":"Metacognition","score":0.6588000059127808},{"id":"https://openalex.org/keywords/control","display_name":"Control (management)","score":0.5938000082969666},{"id":"https://openalex.org/keywords/oracle","display_name":"Oracle","score":0.5899999737739563},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.510699987411499},{"id":"https://openalex.org/keywords/suite","display_name":"Suite","score":0.4837999939918518},{"id":"https://openalex.org/keywords/opportunistic-reasoning","display_name":"Opportunistic reasoning","score":0.44830000400543213},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.44670000672340393}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6692000031471252},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6632000207901001},{"id":"https://openalex.org/C118147538","wikidata":"https://www.wikidata.org/wiki/Q1126970","display_name":"Metacognition","level":3,"score":0.6588000059127808},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.5938000082969666},{"id":"https://openalex.org/C55166926","wikidata":"https://www.wikidata.org/wiki/Q2892946","display_name":"Oracle","level":2,"score":0.5899999737739563},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5476999878883362},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.510699987411499},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.4837999939918518},{"id":"https://openalex.org/C86827895","wikidata":"https://www.wikidata.org/wiki/Q7098582","display_name":"Opportunistic reasoning","level":4,"score":0.44830000400543213},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.44670000672340393},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4406000077724457},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.43459999561309814},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.39800000190734863},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.37880000472068787},{"id":"https://openalex.org/C159032336","wikidata":"https://www.wikidata.org/wiki/Q2488768","display_name":"Non-monotonic logic","level":2,"score":0.37450000643730164},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.3483000099658966},{"id":"https://openalex.org/C89288958","wikidata":"https://www.wikidata.org/wiki/Q7301504","display_name":"Reasoning system","level":2,"score":0.33739998936653137},{"id":"https://openalex.org/C43971567","wikidata":"https://www.wikidata.org/wiki/Q3142865","display_name":"Logical reasoning","level":2,"score":0.3192000091075897},{"id":"https://openalex.org/C36964233","wikidata":"https://www.wikidata.org/wiki/Q7920942","display_name":"Verbal reasoning","level":3,"score":0.3131999969482422},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.28439998626708984},{"id":"https://openalex.org/C97364631","wikidata":"https://www.wikidata.org/wiki/Q484284","display_name":"Deductive reasoning","level":2,"score":0.2703999876976013},{"id":"https://openalex.org/C2780203653","wikidata":"https://www.wikidata.org/wiki/Q17009571","display_name":"Dual process theory (moral psychology)","level":3,"score":0.26899999380111694},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.26649999618530273}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.28135","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.28135","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.28135","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.28135","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"test-time":[1,211],"reasoning":[2,15,43,55,73,212],"methods":[3],"improve":[4],"performance":[5],"by":[6],"generating":[7],"more":[8],"candidate":[9],"chains":[10],"or":[11],"searching":[12],"over":[13,23,53,137],"larger":[14],"trees,":[16],"but":[17],"they":[18],"typically":[19],"lack":[20],"explicit":[21,199],"control":[22,52,201],"when":[24,34],"to":[25,28,31,35],"expand,":[26],"what":[27],"prune,":[29],"how":[30],"repair,":[32,84],"and":[33,75,86,99,130,148,167,172,189,209],"abstain.":[36],"We":[37],"introduce":[38],"CoT2-Meta,":[39],"a":[40,76,161,203],"training-free":[41],"metacognitive":[42,200],"framework":[44,58,157],"that":[45,78,198],"combines":[46],"object-level":[47],"chain-of-thought":[48],"generation":[49],"with":[50,135],"meta-level":[51],"partial":[54],"trajectories.":[56],"The":[57],"integrates":[59],"four":[60],"components:":[61],"strategy-conditioned":[62],"thought":[63],"generation,":[64],"tree-structured":[65],"search,":[66],"an":[67],"online":[68],"process":[69],"oracle":[70],"for":[71,207],"step-level":[72],"evaluation,":[74],"meta-controller":[77],"allocates":[79],"computation":[80],"through":[81],"expansion,":[82],"pruning,":[83],"stopping,":[85],"fallback":[87],"decisions.":[88],"Under":[89],"matched":[90],"inference":[91],"budgets,":[92],"CoT2-Meta":[93],"consistently":[94],"outperforms":[95],"strong":[96],"single-path,":[97],"sampling-based,":[98],"search-based":[100],"baselines,":[101],"including":[102],"ReST-MCTS.":[103],"On":[104],"the":[105,138,156],"default":[106],"backbone,":[107],"it":[108],"achieves":[109],"92.8":[110],"EM":[111,119],"on":[112,116,120,124,128,133],"MATH,":[113],"90.4":[114],"accuracy":[115,123,127,132],"GPQA,":[117],"98.65":[118],"GSM8K,":[121],"75.8":[122],"BBEH,":[125],"85.6":[126],"MMMU-Pro,":[129],"48.8":[131],"HLE,":[134],"gains":[136,191],"strongest":[139],"non-CoT2-Meta":[140],"baseline":[141],"of":[142],"+3.6,":[143],"+5.2,":[144],"+1.15,":[145],"+2.0,":[146],"+4.3,":[147],"+4.3":[149],"points,":[150],"respectively.":[151],"Beyond":[152],"these":[153],"core":[154],"results,":[155],"remains":[158],"effective":[159],"across":[160,192],"broader":[162],"15-benchmark":[163],"suite":[164],"spanning":[165],"knowledge":[166],"QA,":[168],"multi-hop":[169],"reasoning,":[170],"coding,":[171],"out-of-distribution":[173],"evaluation.":[174],"Additional":[175],"analyses":[176],"show":[177],"better":[178],"compute":[179],"scaling,":[180],"improved":[181],"calibration,":[182],"stronger":[183],"selective":[184],"prediction,":[185],"targeted":[186],"repair":[187],"behavior,":[188],"consistent":[190],"backbone":[193],"families.":[194],"These":[195],"results":[196],"suggest":[197],"is":[202],"practical":[204],"design":[205],"principle":[206],"reliable":[208],"compute-efficient":[210],"systems.":[213]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-04-02T00:00:00"}