{"id":"https://openalex.org/W7163177396","doi":"https://doi.org/10.48550/arxiv.2606.01160","title":"Expected Value Alignment for Generative Reward Modeling in Formal Mathematics Verification","display_name":"Expected Value Alignment for Generative Reward Modeling in Formal Mathematics Verification","publication_year":2026,"publication_date":"2026-05-31","ids":{"openalex":"https://openalex.org/W7163177396","doi":"https://doi.org/10.48550/arxiv.2606.01160"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.01160","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01160","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":null,"license_id":null,"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.2606.01160","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137711197","display_name":"Shihao Ji","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji, Shihao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137694149","display_name":"Haotao Tan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tan, Haotao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137624077","display_name":"Zihui Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Zihui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137650172","display_name":"Mingyu Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Mingyu","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/T11948","display_name":"Machine Learning in Materials Science","score":0.17059999704360962,"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":0.17059999704360962,"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/T13523","display_name":"Mathematics, Computing, and Information Processing","score":0.16009999811649323,"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/T10028","display_name":"Topic Modeling","score":0.06800000369548798,"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/interpretability","display_name":"Interpretability","score":0.8604000210762024},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.5999000072479248},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.48570001125335693},{"id":"https://openalex.org/keywords/syntax","display_name":"Syntax","score":0.40799999237060547},{"id":"https://openalex.org/keywords/expected-value","display_name":"Expected value","score":0.400299996137619},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.3912999927997589},{"id":"https://openalex.org/keywords/value","display_name":"Value (mathematics)","score":0.3521000146865845},{"id":"https://openalex.org/keywords/measure","display_name":"Measure (data warehouse)","score":0.3418999910354614},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.3260999917984009}],"concepts":[{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.8604000210762024},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6671000123023987},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.5999000072479248},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5562999844551086},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.49959999322891235},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.48570001125335693},{"id":"https://openalex.org/C60048249","wikidata":"https://www.wikidata.org/wiki/Q37437","display_name":"Syntax","level":2,"score":0.40799999237060547},{"id":"https://openalex.org/C141042865","wikidata":"https://www.wikidata.org/wiki/Q200125","display_name":"Expected value","level":2,"score":0.400299996137619},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.3912999927997589},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.3521000146865845},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.3418999910354614},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3260999917984009},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3257000148296356},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.32170000672340393},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3100999891757965},{"id":"https://openalex.org/C83546350","wikidata":"https://www.wikidata.org/wiki/Q1139051","display_name":"Regression","level":2,"score":0.3021000027656555},{"id":"https://openalex.org/C97137487","wikidata":"https://www.wikidata.org/wiki/Q729138","display_name":"Integer (computer science)","level":2,"score":0.2874000072479248},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2766999900341034},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.2721000015735626},{"id":"https://openalex.org/C2778565505","wikidata":"https://www.wikidata.org/wiki/Q2207566","display_name":"Spec#","level":2,"score":0.2660999894142151},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C139945424","wikidata":"https://www.wikidata.org/wiki/Q1940696","display_name":"Mean squared error","level":2,"score":0.2606000006198883},{"id":"https://openalex.org/C152877465","wikidata":"https://www.wikidata.org/wiki/Q208042","display_name":"Regression analysis","level":2,"score":0.25999999046325684},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2587999999523163},{"id":"https://openalex.org/C22019652","wikidata":"https://www.wikidata.org/wiki/Q331309","display_name":"Overfitting","level":3,"score":0.257999986410141},{"id":"https://openalex.org/C142998047","wikidata":"https://www.wikidata.org/wiki/Q193983","display_name":"Metalanguage","level":2,"score":0.25209999084472656},{"id":"https://openalex.org/C111498074","wikidata":"https://www.wikidata.org/wiki/Q173326","display_name":"Formal verification","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.01160","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01160","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.01160","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01160","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":null,"license_id":null,"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":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"are":[4,62,72],"increasingly":[5],"used":[6],"with":[7,19,134],"formal":[8,156],"interactive":[9],"theorem":[10],"provers":[11],"such":[12],"as":[13,116],"Lean":[14,154],"4.":[15],"Scaling":[16],"these":[17,142],"systems":[18],"reinforcement":[20],"learning":[21],"or":[22],"search":[23],"methods":[24],"requires":[25],"process":[26],"reward":[27,56,151],"models":[28,44,57],"(PRMs)":[29],"that":[30,85,169],"can":[31],"evaluate":[32,159],"intermediate":[33],"reasoning":[34],"steps.":[35],"Existing":[36],"reward-model":[37],"designs":[38],"expose":[39],"a":[40,82,106,113,150],"practical":[41],"trade-off.":[42],"Value-head":[43],"provide":[45],"continuous":[46,66,93,114,170],"scores":[47,94,104],"but":[48,61],"modify":[49],"the":[50,87,96,117,120,123,129,179],"generative":[51,55,182],"model":[52,101,152],"interface,":[53],"while":[54,91,177],"preserve":[58],"textual":[59],"rationales":[60],"poorly":[63],"matched":[64],"to":[65],"floating-point":[67],"regression":[68],"because":[69],"numeric":[70],"values":[71],"split":[73],"across":[74],"tokens.":[75,126],"We":[76,145],"introduce":[77],"Expected":[78],"Value":[79],"Alignment":[80],"(EVA),":[81],"reward-modeling":[83,164],"procedure":[84],"keeps":[86],"surface":[88],"output":[89],"discrete":[90],"extracting":[92],"from":[95],"model's":[97],"token":[98],"distribution.":[99],"The":[100,166],"emits":[102],"integer":[103],"in":[105,148],"structured":[107],"JSON":[108],"format,":[109],"and":[110,158,163],"EVA":[111,147],"computes":[112],"score":[115],"expectation":[118],"over":[119],"logits":[121],"of":[122,181],"corresponding":[124],"anchor":[125],"Training":[127],"combines":[128],"causal":[130],"language":[131],"modeling":[132],"objective":[133],"an":[135],"auxiliary":[136],"mean":[137],"squared":[138],"error":[139],"loss":[140],"on":[141],"expected":[143],"values.":[144],"instantiate":[146],"\\textit{Leibniz},":[149],"for":[153],"4":[155],"verification,":[157],"it":[160],"against":[161],"zero-shot":[162],"baselines.":[165],"evaluation":[167],"demonstrates":[168],"logit-based":[171],"scoring":[172],"significantly":[173],"reduces":[174],"discretization":[175],"artifacts":[176],"retaining":[178],"interpretability":[180],"critiques.":[183]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-03T00:00:00"}