{"id":"https://openalex.org/W7151409912","doi":"https://doi.org/10.48550/arxiv.2604.04898","title":"QED-Nano: Teaching a Tiny Model to Prove Hard Theorems","display_name":"QED-Nano: Teaching a Tiny Model to Prove Hard Theorems","publication_year":2026,"publication_date":"2026-04-06","ids":{"openalex":"https://openalex.org/W7151409912","doi":"https://doi.org/10.48550/arxiv.2604.04898"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.04898","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.04898","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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.04898","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133126676","display_name":"LM-Provers","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"LM-Provers","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133072892","display_name":"Yuxiao Qu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qu, Yuxiao","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133067655","display_name":"Amrith Setlur","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Setlur, Amrith","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5093356429","display_name":"Jasper Dekoninck","orcid":"https://orcid.org/0009-0009-0621-170X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dekoninck, Jasper","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013995783","display_name":"Edward Beeching","orcid":"https://orcid.org/0000-0003-1165-3464"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Beeching, Edward","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133086778","display_name":"Jia Wei LI","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Jia","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133124839","display_name":"Ian Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Ian","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019163452","display_name":"Lewis Tunstall","orcid":"https://orcid.org/0000-0003-3309-5520"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tunstall, Lewis","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5102493293","display_name":"Aviral Kumar","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kumar, Aviral","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"corresponding_author_ids":["https://openalex.org/A5133126676"],"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/T11948","display_name":"Machine Learning in Materials Science","score":0.12269999831914902,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11948","display_name":"Machine Learning in Materials Science","score":0.12269999831914902,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11596","display_name":"Constraint Satisfaction and Optimization","score":0.07129999995231628,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T10126","display_name":"Logic, programming, and type systems","score":0.06129999831318855,"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/mathematical-proof","display_name":"Mathematical proof","score":0.59579998254776},{"id":"https://openalex.org/keywords/fraction","display_name":"Fraction (chemistry)","score":0.47850000858306885},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4578000009059906},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.4284000098705292},{"id":"https://openalex.org/keywords/olympiad","display_name":"Olympiad","score":0.3562000095844269},{"id":"https://openalex.org/keywords/variety","display_name":"Variety (cybernetics)","score":0.35499998927116394}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6391000151634216},{"id":"https://openalex.org/C108710211","wikidata":"https://www.wikidata.org/wiki/Q11538","display_name":"Mathematical proof","level":2,"score":0.59579998254776},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4862000048160553},{"id":"https://openalex.org/C149629883","wikidata":"https://www.wikidata.org/wiki/Q660926","display_name":"Fraction (chemistry)","level":2,"score":0.47850000858306885},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4578000009059906},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4284000098705292},{"id":"https://openalex.org/C130383907","wikidata":"https://www.wikidata.org/wiki/Q221956","display_name":"Olympiad","level":2,"score":0.3562000095844269},{"id":"https://openalex.org/C136197465","wikidata":"https://www.wikidata.org/wiki/Q1729295","display_name":"Variety (cybernetics)","level":2,"score":0.35499998927116394},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.32829999923706055},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.326200008392334},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.32019999623298645},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.3091999888420105},{"id":"https://openalex.org/C2778671685","wikidata":"https://www.wikidata.org/wiki/Q219239","display_name":"Recipe","level":2,"score":0.30799999833106995},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.28040000796318054},{"id":"https://openalex.org/C76969082","wikidata":"https://www.wikidata.org/wiki/Q486902","display_name":"Mathematical model","level":2,"score":0.2623000144958496},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2590000033378601}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.04898","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.04898","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":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.04898","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.04898","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":"article"},"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":{"Proprietary":[0],"AI":[1],"systems":[2,29],"have":[3],"recently":[4],"demonstrated":[5],"impressive":[6],"capabilities":[7],"on":[8,36,74,177],"complex":[9],"proof-based":[10],"problems,":[11],"with":[12,117,124],"gold-level":[13],"performance":[14,73,145,158],"reported":[15],"at":[16,166],"the":[17,24,143,157,170,183,188,193,199],"2025":[18],"International":[19],"Mathematical":[20],"Olympiad":[21],"(IMO).":[22],"However,":[23],"training":[25,96,200],"pipelines":[26],"behind":[27],"these":[28],"remain":[30],"largely":[31],"undisclosed,":[32],"and":[33,40,50,120,136,153,155,190,195,198,201],"their":[34],"reliance":[35],"large":[37],"\"internal\"":[38],"models":[39,65,161],"scaffolds":[41],"makes":[42],"them":[43],"expensive":[44],"to":[45,48,52,69,104],"run,":[46],"difficult":[47,75],"reproduce,":[49],"hard":[51],"study":[53],"or":[54],"improve":[55],"upon.":[56],"This":[57],"raises":[58],"a":[59,88,125,167],"central":[60],"question:":[61],"can":[62],"small,":[63],"open":[64,149,178],"also":[66],"be":[67],"trained":[68],"achieve":[70],"competitive":[71],"reasoning":[72,126],"Olympiad-level":[76,93],"math?":[77],"In":[78],"this":[79,83],"paper,":[80],"we":[81,181],"answer":[82],"question":[84],"by":[85,109],"building":[86],"QED-Nano,":[87],"4B":[89],"model":[90],"post-trained":[91],"for":[92],"proofs.":[94],"Our":[95],"recipe":[97],"has":[98],"three":[99],"stages:":[100],"(1)":[101],"supervised":[102],"fine-tuning":[103],"imbue":[105],"good":[106],"proof-writing":[107],"styles":[108],"distilling":[110],"from":[111],"DeepSeek-Math-V2,":[112],"(2)":[113],"reinforcement":[114],"learning":[115],"(RL)":[116],"rubric-based":[118],"rewards,":[119],"(3)":[121],"expanding":[122],"RL":[123],"cache,":[127],"which":[128],"decomposes":[129],"long":[130],"proofs":[131],"into":[132],"iterative":[133],"summarize-and-refine":[134],"cycles":[135],"enables":[137],"stronger":[138],"test-time":[139],"reasoning.":[140],"QED-Nano":[141,185,189],"surpasses":[142],"proof-generation":[144],"of":[146,159,169],"much":[147],"larger":[148],"models,":[150,192],"including":[151,187],"Nomos-1":[152],"GPT-OSS-120B,":[154],"approaches":[156],"proprietary":[160],"like":[162],"Gemini":[163],"3":[164],"Pro,":[165],"fraction":[168],"inference":[171],"cost.":[172],"To":[173],"support":[174],"further":[175],"research":[176],"mathematical":[179],"reasoning,":[180],"release":[182],"full":[184],"pipeline,":[186],"QED-Nano-SFT":[191],"FineProofs-SFT":[194],"FineProofs-RL":[196],"datasets,":[197],"evaluation":[202],"code.":[203]},"counts_by_year":[],"updated_date":"2026-04-08T06:07:18.267832","created_date":"2026-04-08T00:00:00"}