{"id":"https://openalex.org/W7167619593","doi":"https://doi.org/10.48550/arxiv.2607.05391","title":"LLM-as-a-Verifier: A General-Purpose Verification Framework","display_name":"LLM-as-a-Verifier: A General-Purpose Verification Framework","publication_year":2026,"publication_date":"2026-07-06","ids":{"openalex":"https://openalex.org/W7167619593","doi":"https://doi.org/10.48550/arxiv.2607.05391"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2607.05391","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.05391","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,"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.2607.05391","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5140207543","display_name":"Jacky Kwok","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kwok, Jacky","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019366488","display_name":"Shulu Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Shulu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033692685","display_name":"Pranav Atreya","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Atreya, Pranav","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140197742","display_name":"Yuejiang Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Yuejiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140202399","display_name":"Yixing Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiang, Yixing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140211724","display_name":"Chelsea Finn","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Finn, Chelsea","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140162241","display_name":"Marco Pavone","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pavone, Marco","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140161901","display_name":"Ion Stoica","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Stoica, Ion","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5140159351","display_name":"Azalia Mirhoseini","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mirhoseini, Azalia","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/T10653","display_name":"Robot Manipulation and Learning","score":0.10189999639987946,"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"}},"topics":[{"id":"https://openalex.org/T10653","display_name":"Robot Manipulation and Learning","score":0.10189999639987946,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.060600001364946365,"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/T11574","display_name":"Artificial Intelligence in Games","score":0.06030000001192093,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/correctness","display_name":"Correctness","score":0.7594000101089478},{"id":"https://openalex.org/keywords/probabilistic-logic","display_name":"Probabilistic logic","score":0.6078000068664551},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5403000116348267},{"id":"https://openalex.org/keywords/ranking","display_name":"Ranking (information retrieval)","score":0.5321000218391418},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.5144000053405762},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.4366999864578247},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.42170000076293945},{"id":"https://openalex.org/keywords/sample","display_name":"Sample (material)","score":0.35109999775886536}],"concepts":[{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.7594000101089478},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6998000144958496},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.6078000068664551},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5403000116348267},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.5321000218391418},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.5144000053405762},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.49480000138282776},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.45509999990463257},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.4366999864578247},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.42170000076293945},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.40209999680519104},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.35109999775886536},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.3375999927520752},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.32659998536109924},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.3043999969959259},{"id":"https://openalex.org/C129848803","wikidata":"https://www.wikidata.org/wiki/Q2564360","display_name":"Sample size determination","level":2,"score":0.2921999990940094},{"id":"https://openalex.org/C2778029271","wikidata":"https://www.wikidata.org/wiki/Q5421931","display_name":"Extension (predicate logic)","level":2,"score":0.2906000018119812},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.28519999980926514},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.28130000829696655},{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.2694000005722046},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.26750001311302185},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.265500009059906},{"id":"https://openalex.org/C34413123","wikidata":"https://www.wikidata.org/wiki/Q170978","display_name":"Robotics","level":3,"score":0.2515999972820282}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2607.05391","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.05391","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,"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.2607.05391","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2607.05391","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,"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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Scaling":[0],"pre-training,":[1],"post-training,":[2],"and":[3,40,107,126,138,151,186,217,240,244],"test-time":[4],"compute":[5],"have":[6],"become":[7],"the":[8,13,23,27,78,81,117,163,169,191,235],"central":[9],"paradigms":[10],"for":[11,55,73,161,202,210,232],"improving":[12,234],"capabilities":[14],"of":[15,29,83,238],"LLMs.":[16],"In":[17,111],"this":[18,39],"work,":[19],"we":[20,44,113,224],"identify":[21],"verification,":[22,190],"ability":[24],"to":[25,69,87,96,121,143,215],"determine":[26],"correctness":[28],"a":[30,33,47,157,200],"solution,":[31],"as":[32,199],"new":[34],"scaling":[35,116,135],"axis.":[36],"To":[37],"unlock":[38],"demonstrate":[41],"its":[42],"effectiveness,":[43],"introduce":[45,156],"LLM-as-a-Verifier,":[46],"general-purpose":[48],"verification":[49,95,147],"framework":[50],"that":[51,66,115,226],"provides":[52],"fine-grained":[53,192],"feedback":[54,231],"agentic":[56,221],"tasks":[57],"without":[58],"requiring":[59],"additional":[60,144],"training.":[61],"Unlike":[62],"standard":[63],"LM":[64],"judges":[65],"prompt":[67],"LLMs":[68],"produce":[70],"discrete":[71],"scores":[72],"candidate":[74],"solutions,":[75,128],"LLM-as-a-Verifier":[76,173,195,227],"computes":[77],"expectation":[79],"over":[80],"distribution":[82],"scoring":[84,118],"token":[85],"logits":[86],"generate":[88],"continuous":[89,171],"scores.":[90,172],"This":[91],"probabilistic":[92],"formulation":[93],"enables":[94],"scale":[97],"along":[98],"multiple":[99],"dimensions:":[100],"(1)":[101],"score":[102],"granularity,":[103],"(2)":[104],"repeated":[105,136],"evaluation,":[106],"(3)":[108],"criteria":[109,139],"decomposition.":[110],"particular,":[112],"show":[114,225],"granularity":[119],"leads":[120],"better":[122],"separation":[123],"between":[124],"positive":[125],"negative":[127],"resulting":[129],"in":[130,146],"more":[131],"calibrated":[132],"comparisons.":[133],"Moreover,":[134],"evaluation":[137],"decomposition":[140],"consistently":[141],"lead":[142],"gains":[145],"accuracy":[148],"through":[149],"variance":[150],"complexity":[152],"reduction.":[153],"We":[154,206],"further":[155],"cost-efficient":[158],"ranking":[159],"algorithm":[160],"selecting":[162],"best":[164],"solution":[165],"among":[166],"candidates":[167],"using":[168],"verifier's":[170],"achieves":[174],"state-of-the-art":[175],"performance":[176],"on":[177,242],"Terminal-Bench":[178],"V2":[179],"(86.5%),":[180],"SWE-Bench":[181],"Verified":[182],"(78.2%),":[183],"RoboRewardBench":[184],"(87.4%),":[185],"MedAgentBench":[187],"(73.3%).":[188],"Beyond":[189],"signals":[193],"from":[194],"can":[196,228],"also":[197],"serve":[198],"proxy":[201],"estimating":[203],"task":[204],"progress.":[205],"build":[207],"an":[208],"extension":[209],"Claude":[211],"Code,":[212],"enabling":[213],"developers":[214],"monitor":[216],"improve":[218],"their":[219],"own":[220],"systems.":[222],"Finally,":[223],"provide":[229],"dense":[230],"RL,":[233],"sample":[236],"efficiency":[237],"SAC":[239],"GRPO":[241],"robotics":[243],"mathematical":[245],"reasoning":[246],"benchmarks.":[247]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-07-08T00:00:00"}