{"id":"https://openalex.org/W4407123355","doi":"https://doi.org/10.48550/arxiv.2502.00271","title":"Scaling Flaws of Verifier-Guided Search in Mathematical Reasoning","display_name":"Scaling Flaws of Verifier-Guided Search in Mathematical Reasoning","publication_year":2025,"publication_date":"2025-02-01","ids":{"openalex":"https://openalex.org/W4407123355","doi":"https://doi.org/10.48550/arxiv.2502.00271"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2502.00271","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2502.00271","pdf_url":"https://arxiv.org/pdf/2502.00271","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"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2502.00271","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5103244124","display_name":"Fei Yu","orcid":"https://orcid.org/0000-0003-0193-1790","observed_orcids":["https://orcid.org/0000-0003-0193-1790","https://orcid.org/0000-0001-8970-6269"]},"institutions":[{"id":"https://openalex.org/I4210116924","display_name":"Chinese University of Hong Kong, Shenzhen","ror":"https://ror.org/02d5ks197","country_code":"CN","type":"education","lineage":["https://openalex.org/I177725633","https://openalex.org/I180726961","https://openalex.org/I4210116924"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yu, Fei","raw_affiliation_strings":["The Chinese University of Hong Kong , Shenzhen , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong , Shenzhen , China","institution_ids":["https://openalex.org/I4210116924"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024774792","display_name":"Yingru Li","orcid":"https://orcid.org/0000-0002-3258-9230","observed_orcids":["https://orcid.org/0000-0002-3258-9230","https://orcid.org/0009-0007-8218-4649"]},"institutions":[{"id":"https://openalex.org/I4210116924","display_name":"Chinese University of Hong Kong, Shenzhen","ror":"https://ror.org/02d5ks197","country_code":"CN","type":"education","lineage":["https://openalex.org/I177725633","https://openalex.org/I180726961","https://openalex.org/I4210116924"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Li, Yingru","raw_affiliation_strings":["The Chinese University of Hong Kong , Shenzhen , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong , Shenzhen , China","institution_ids":["https://openalex.org/I4210116924"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5057282504","display_name":"Benyou Wang","orcid":"https://orcid.org/0000-0002-1501-9914","observed_orcids":["https://orcid.org/0000-0002-1501-9914","https://orcid.org/0009-0000-3606-1364"]},"institutions":[{"id":"https://openalex.org/I4210116924","display_name":"Chinese University of Hong Kong, Shenzhen","ror":"https://ror.org/02d5ks197","country_code":"CN","type":"education","lineage":["https://openalex.org/I177725633","https://openalex.org/I180726961","https://openalex.org/I4210116924"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wang, Benyou","raw_affiliation_strings":["The Chinese University of Hong Kong , Shenzhen , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong , Shenzhen , China","institution_ids":["https://openalex.org/I4210116924"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I4210116924"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":16,"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":0.42329999804496765,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"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":0.42329999804496765,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"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":0.10989999771118164,"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":0.10379999876022339,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"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/multi-step-reasoning","display_name":"multi-step reasoning","score":0.6769999861717224},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8180000185966492},{"id":"https://openalex.org/keywords/inference-time-scaling","display_name":"inference-time scaling","score":0.828000009059906},{"id":"https://openalex.org/keywords/process-reward-models","display_name":"process reward models","score":0.6190000176429749},{"id":"https://openalex.org/keywords/mathematical-reasoning","display_name":"mathematical reasoning","score":0.7070000171661377}],"concepts":[{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.7158272862434387},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6044119000434875},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.36306294798851013},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.1868564784526825},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2502.00271","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2502.00271","pdf_url":"https://arxiv.org/pdf/2502.00271","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"},{"id":"doi:10.48550/arxiv.2502.00271","doi":"https://doi.org/10.48550/arxiv.2502.00271","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2502.00271","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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2502.00271","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2502.00271","pdf_url":"https://arxiv.org/pdf/2502.00271","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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4407123355.pdf","grobid_xml":"https://content.openalex.org/works/W4407123355.grobid-xml"},"referenced_works_count":14,"referenced_works":["https://openalex.org/W3177813494","https://openalex.org/W3210904070","https://openalex.org/W4387321091","https://openalex.org/W4387561528","https://openalex.org/W4389520747","https://openalex.org/W4391631327","https://openalex.org/W4401042684","https://openalex.org/W4401307656","https://openalex.org/W4402671689","https://openalex.org/W4402733601","https://openalex.org/W4406023812","https://openalex.org/W4406492615","https://openalex.org/W4415797522","https://openalex.org/W4415798274"],"related_works":["https://openalex.org/W4403346096","https://openalex.org/W4403928622","https://openalex.org/W4414750860","https://openalex.org/W7213268218","https://openalex.org/W4416361203","https://openalex.org/W4403578379","https://openalex.org/W4403570105","https://openalex.org/W4415277368","https://openalex.org/W4416034322","https://openalex.org/W7171718126"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2,48,62],"(LLMs)":[3],"struggle":[4],"with":[5],"multi-step":[6],"reasoning,":[7],"where":[8,88],"inference-time":[9],"scaling":[10,43],"has":[11],"emerged":[12],"as":[13],"a":[14,40],"promising":[15],"strategy":[16],"for":[17],"performance":[18],"improvement.":[19],"Verifier-guided":[20],"search":[21,72,110,137],"outperforms":[22],"repeated":[23,79],"sampling":[24],"when":[25],"sample":[26,68],"size":[27,69],"is":[28],"limited":[29],"by":[30],"selecting":[31],"and":[32,51,56,58,63,76,93,106,122,138],"prioritizing":[33],"valid":[34,97],"reasoning":[35],"paths.":[36,98],"However,":[37],"we":[38,116],"identify":[39],"critical":[41],"limitation:":[42],"flaws,":[44],"prevalent":[45],"across":[46],"different":[47],"(Mistral":[49],"7B":[50],"DeepSeekMath":[52],"7B),":[53],"benchmarks":[54],"(GSM8K":[55],"MATH),":[57],"verifiers":[59,90,121],"(outcome":[60],"value":[61],"process":[64],"reward":[65],"models).":[66],"As":[67],"increases,":[70],"verifier-guided":[71,136],"exhibits":[73],"diminishing":[74],"advantages":[75],"eventually":[77],"underperforms":[78],"sampling.":[80],"Our":[81,130],"analysis":[82],"attributes":[83],"this":[84],"to":[85],"verifier":[86,114],"failures,":[87,115],"imperfect":[89],"misrank":[91],"candidates":[92],"erroneously":[94],"prune":[95],"all":[96],"These":[99],"issues":[100],"are":[101],"further":[102],"exacerbated":[103],"in":[104,135],"challenging":[105],"out-of-distribution":[107],"problems,":[108],"restricting":[109],"effectiveness.":[111],"To":[112],"mitigate":[113],"explore":[117],"reducing":[118],"reliance":[119],"on":[120],"conduct":[123],"preliminary":[124],"investigations":[125],"using":[126],"two":[127],"simple":[128],"methods.":[129],"findings":[131],"reveal":[132],"fundamental":[133],"limitations":[134],"suggest":[139],"future":[140],"directions.":[141]},"counts_by_year":[{"year":2026,"cited_by_count":11},{"year":2025,"cited_by_count":5}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-02-05T00:00:00"}