{"id":"https://openalex.org/W4417488762","doi":"https://doi.org/10.48550/arxiv.2507.15225","title":"Solving Formal Math Problems by Decomposition and Iterative Reflection","display_name":"Solving Formal Math Problems by Decomposition and Iterative Reflection","publication_year":2025,"publication_date":"2025-07-21","ids":{"openalex":"https://openalex.org/W4417488762","doi":"https://doi.org/10.48550/arxiv.2507.15225"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2507.15225","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2507.15225","pdf_url":"https://arxiv.org/pdf/2507.15225","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"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2507.15225","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Zhou, Yichi","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Yichi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101154136","display_name":"Jianqiu Zhao","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Jianqiu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5039281149","display_name":"Yongxin Zhang","orcid":"https://orcid.org/0000-0003-1525-5893","observed_orcids":["https://orcid.org/0000-0003-1525-5893"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Yongxin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002671826","display_name":"Bohan Wang","orcid":"https://orcid.org/0009-0005-6490-3203","observed_orcids":["https://orcid.org/0009-0005-6490-3203"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Bohan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101657703","display_name":"Si Ran Wang","orcid":"https://orcid.org/0000-0002-8937-9031","observed_orcids":["https://orcid.org/0000-0002-8937-9031"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Siran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044778112","display_name":"Luoxin Chen","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Luoxin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100446916","display_name":"Jiahui Wang","orcid":"https://orcid.org/0000-0002-6438-1506","observed_orcids":["https://orcid.org/0000-0002-6438-1506","https://orcid.org/0009-0005-9327-7096"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Jiahui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100772080","display_name":"Haowei Chen","orcid":"https://orcid.org/0000-0003-3429-9146","observed_orcids":["https://orcid.org/0000-0003-3429-9146","https://orcid.org/0009-0005-8495-290X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Haowei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120854318","display_name":"Allan Jie","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jie, Allan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5087936668","display_name":"Xinbo Zhang","orcid":"https://orcid.org/0000-0002-5806-159X","observed_orcids":["https://orcid.org/0000-0002-5806-159X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Xinbo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102921879","display_name":"Haocheng Wang","orcid":"https://orcid.org/0000-0001-8078-9475","observed_orcids":["https://orcid.org/0000-0001-8078-9475"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Haocheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107403608","display_name":"Luong Trung","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Trung, Luong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103167214","display_name":"Rong Ye","orcid":"https://orcid.org/0009-0007-0690-0535","observed_orcids":["https://orcid.org/0009-0007-0690-0535"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ye, Rong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5077371257","display_name":"Phi-Hung Hoang","orcid":"https://orcid.org/0009-0002-2967-5952","observed_orcids":["https://orcid.org/0009-0002-2967-5952"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hoang, Phan Nhat","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066928490","display_name":"Huishuai Zhang","orcid":"https://orcid.org/0000-0003-2711-7295","observed_orcids":["https://orcid.org/0000-0003-2711-7295"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Huishuai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041249182","display_name":"Peng Sun","orcid":"https://orcid.org/0009-0004-1484-2660","observed_orcids":["https://orcid.org/0009-0004-1484-2660","https://orcid.org/0000-0003-4424-5969"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Peng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Li, Hang","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Hang","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/T10126","display_name":"Logic, programming, and type systems","score":0.5785999894142151,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"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":0.5785999894142151,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"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":0.25459998846054077,"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":0.033900000154972076,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"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/lean-4","display_name":"Lean 4","score":0.9810000061988831},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8149999976158142},{"id":"https://openalex.org/keywords/llm-based-agents","display_name":"LLM agents","score":0.7020000219345093},{"id":"https://openalex.org/keywords/theorem-proving","display_name":"theorem proving","score":0.765999972820282},{"id":"https://openalex.org/keywords/domain-specific-languages","display_name":"domain-specific languages","score":0.8240000009536743},{"id":"https://openalex.org/keywords/automated-verification","display_name":"automated verification","score":0.4749999940395355},{"id":"https://openalex.org/keywords/test-time-scaling","display_name":"test-time scaling","score":0.5419999957084656}],"concepts":[{"id":"https://openalex.org/C108710211","wikidata":"https://www.wikidata.org/wiki/Q11538","display_name":"Mathematical proof","level":2,"score":0.7836999893188477},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6234999895095825},{"id":"https://openalex.org/C159718280","wikidata":"https://www.wikidata.org/wiki/Q5526353","display_name":"Gas meter prover","level":3,"score":0.6197999715805054},{"id":"https://openalex.org/C39920170","wikidata":"https://www.wikidata.org/wiki/Q693083","display_name":"Soundness","level":2,"score":0.6029999852180481},{"id":"https://openalex.org/C206880738","wikidata":"https://www.wikidata.org/wiki/Q431667","display_name":"Automated theorem proving","level":2,"score":0.5629000067710876},{"id":"https://openalex.org/C94461902","wikidata":"https://www.wikidata.org/wiki/Q2762418","display_name":"Formal proof","level":3,"score":0.4733000099658966},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.4690999984741211},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.4505000114440918},{"id":"https://openalex.org/C65682993","wikidata":"https://www.wikidata.org/wiki/Q1056451","display_name":"Reflection (computer programming)","level":2,"score":0.4503999948501587},{"id":"https://openalex.org/C75606506","wikidata":"https://www.wikidata.org/wiki/Q1049183","display_name":"Formal methods","level":2,"score":0.4433000087738037},{"id":"https://openalex.org/C111498074","wikidata":"https://www.wikidata.org/wiki/Q173326","display_name":"Formal verification","level":2,"score":0.4415000081062317},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.4381999969482422},{"id":"https://openalex.org/C179518139","wikidata":"https://www.wikidata.org/wiki/Q5140297","display_name":"Coding (social sciences)","level":2,"score":0.41519999504089355},{"id":"https://openalex.org/C203265346","wikidata":"https://www.wikidata.org/wiki/Q11387554","display_name":"Proof assistant","level":3,"score":0.4092000126838684},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.40709999203681946},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.39899998903274536},{"id":"https://openalex.org/C124681953","wikidata":"https://www.wikidata.org/wiki/Q339062","display_name":"Decomposition","level":2,"score":0.3846000134944916},{"id":"https://openalex.org/C116253237","wikidata":"https://www.wikidata.org/wiki/Q1437424","display_name":"Formal specification","level":2,"score":0.33219999074935913},{"id":"https://openalex.org/C185874996","wikidata":"https://www.wikidata.org/wiki/Q269699","display_name":"Interdependence","level":2,"score":0.2924000024795532},{"id":"https://openalex.org/C102315432","wikidata":"https://www.wikidata.org/wiki/Q649732","display_name":"Formal system","level":2,"score":0.28119999170303345},{"id":"https://openalex.org/C2777686260","wikidata":"https://www.wikidata.org/wiki/Q144037","display_name":"Calculus (dental)","level":2,"score":0.28029999136924744},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2777000069618225},{"id":"https://openalex.org/C188198153","wikidata":"https://www.wikidata.org/wiki/Q1613840","display_name":"Limiting","level":2,"score":0.27730000019073486},{"id":"https://openalex.org/C143587482","wikidata":"https://www.wikidata.org/wiki/Q1543216","display_name":"Iterative and incremental development","level":2,"score":0.26339998841285706},{"id":"https://openalex.org/C156325763","wikidata":"https://www.wikidata.org/wiki/Q1930895","display_name":"Operational semantics","level":3,"score":0.26159998774528503},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.25540000200271606},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2529999911785126},{"id":"https://openalex.org/C2318724","wikidata":"https://www.wikidata.org/wiki/Q852732","display_name":"Proof theory","level":3,"score":0.25279998779296875},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.25279998779296875},{"id":"https://openalex.org/C2779982483","wikidata":"https://www.wikidata.org/wiki/Q6094420","display_name":"Iterative refinement","level":2,"score":0.2515000104904175}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2507.15225","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2507.15225","pdf_url":"https://arxiv.org/pdf/2507.15225","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"},{"id":"doi:10.48550/arxiv.2507.15225","doi":"https://doi.org/10.48550/arxiv.2507.15225","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2507.15225","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:2507.15225","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2507.15225","pdf_url":"https://arxiv.org/pdf/2507.15225","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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4415331377","https://openalex.org/W4404783189","https://openalex.org/W7163338413","https://openalex.org/W4415337241","https://openalex.org/W4417525574","https://openalex.org/W4417268860","https://openalex.org/W4398796786","https://openalex.org/W6948275686","https://openalex.org/W4414580377","https://openalex.org/W4417426748"],"abstract_inverted_index":{"General-purpose":[0],"Large":[1],"Language":[2,146],"Models":[3],"(LLMs)":[4],"have":[5],"achieved":[6],"remarkable":[7],"success":[8,162],"in":[9,30,46,113,225],"intelligence,":[10],"performing":[11],"comparably":[12],"to":[13,108,188,218],"human":[14],"experts":[15],"on":[16,61,164],"complex":[17,47],"reasoning":[18,103,224],"tasks":[19],"such":[20],"as":[21],"coding":[22],"and":[23,50,71,91,102,138,142],"mathematical":[24],"reasoning.":[25],"However,":[26],"generating":[27],"formal":[28,63,111,226],"proofs":[29,112],"specialized":[31,219],"languages":[32],"like":[33],"Lean":[34,93,114,150],"4":[35,94,151],"remains":[36],"a":[37,88,143,159,181,214],"significant":[38],"challenge":[39],"for":[40,68,119,135,152,221],"these":[41],"models,":[42],"limiting":[43],"their":[44],"application":[45],"theorem":[48],"proving":[49],"automated":[51,223],"verification.":[52],"Current":[53],"approaches":[54],"typically":[55],"require":[56],"specializing":[57],"models":[58,220],"through":[59],"fine-tuning":[60],"dedicated":[62],"corpora,":[64],"incurring":[65],"high":[66],"costs":[67],"data":[69],"collection":[70],"training.":[72],"In":[73],"this":[74],"work,":[75],"we":[76],"introduce":[77],"\\textbf{Delta":[78,156],"Prover},":[79],"an":[80,132,203],"agent-based":[81],"framework":[82,134],"that":[83,197],"orchestrates":[84],"the":[85,92,100,117,125,165],"interaction":[86],"between":[87],"general-purpose":[89,106,198],"LLM":[90],"proof":[95,140,191],"environment.":[96],"Delta":[97,178],"Prover":[98,157,179],"leverages":[99],"reflection":[101],"capabilities":[104],"of":[105],"LLMs":[107],"interactively":[109],"construct":[110],"4,":[115],"circumventing":[116],"need":[118],"model":[120,175],"specialization.":[121],"At":[122],"its":[123],"core,":[124],"agent":[126],"integrates":[127],"two":[128],"novel,":[129],"interdependent":[130],"components:":[131],"algorithmic":[133],"reflective":[136],"decomposition":[137],"iterative":[139],"repair,":[141],"custom":[144],"Domain-Specific":[145],"(DSL)":[147],"built":[148],"upon":[149],"streamlined":[153],"subproblem":[154],"management.":[155],"achieves":[158],"state-of-the-art":[160],"95.9\\%":[161],"rate":[163],"miniF2F-test":[166],"benchmark,":[167],"surpassing":[168],"all":[169],"existing":[170],"approaches,":[171],"including":[172],"those":[173],"requiring":[174],"specialization.}":[176],"Furthermore,":[177],"exhibits":[180],"significantly":[182],"stronger":[183],"test-time":[184],"scaling":[185],"law":[186],"compared":[187],"standard":[189],"Best-of-N":[190],"strategies.":[192],"Crucially,":[193],"our":[194],"findings":[195],"demonstrate":[196],"LLMs,":[199],"when":[200],"guided":[201],"by":[202],"effective":[204],"agentic":[205],"structure,":[206],"possess":[207],"substantial":[208],"untapped":[209],"theorem-proving":[210],"capabilities.":[211],"This":[212],"presents":[213],"computationally":[215],"efficient":[216],"alternative":[217],"robust":[222],"environments.":[227]},"counts_by_year":[],"updated_date":"2026-10-29T09:56:32.521200","created_date":"2025-10-10T00:00:00"}