{"id":"https://openalex.org/W4416612957","doi":"https://doi.org/10.48550/arxiv.2505.23667","title":"Formula-R1: Incentivizing LLM Reasoning over Complex Tables with Numerical Computation via Formula-Driven Reinforcement Learning","display_name":"Formula-R1: Incentivizing LLM Reasoning over Complex Tables with Numerical Computation via Formula-Driven Reinforcement Learning","publication_year":2025,"publication_date":"2025-05-29","ids":{"openalex":"https://openalex.org/W4416612957","doi":"https://doi.org/10.48550/arxiv.2505.23667"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2505.23667","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2505.23667","pdf_url":"https://arxiv.org/pdf/2505.23667","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","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2505.23667","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5113039527","display_name":"Lang Cao","orcid":"https://orcid.org/0000-0003-0011-5724","observed_orcids":["https://orcid.org/0000-0003-0011-5724"]},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Cao, Lang","raw_affiliation_strings":["University of Illinois Urbana-Champaign"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Illinois Urbana-Champaign","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050168136","display_name":"Jingxian Xu","orcid":"https://orcid.org/0009-0004-1145-6190","observed_orcids":["https://orcid.org/0009-0004-1145-6190"]},"institutions":[{"id":"https://openalex.org/I205237279","display_name":"Nankai University","ror":"https://ror.org/01y1kjr75","country_code":"CN","type":"education","lineage":["https://openalex.org/I205237279"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xu, Jingxian","raw_affiliation_strings":["Nankai University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nankai University","institution_ids":["https://openalex.org/I205237279"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Liu, Hanbing","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Liu, Hanbing","raw_affiliation_strings":["Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100614575","display_name":"Jinyu Wang","orcid":"https://orcid.org/0000-0002-9449-3453","observed_orcids":["https://orcid.org/0000-0002-9449-3453"]},"institutions":[{"id":"https://openalex.org/I1290206253","display_name":"Microsoft (United States)","ror":"https://ror.org/00d0nc645","country_code":"US","type":"company","lineage":["https://openalex.org/I1290206253"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Wang, Jinyu","raw_affiliation_strings":["Shandong University Microsoft Research"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shandong University Microsoft Research","institution_ids":["https://openalex.org/I1290206253"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101603009","display_name":"Mengyu Zhou","orcid":"https://orcid.org/0000-0002-0322-7513","observed_orcids":["https://orcid.org/0000-0002-0322-7513","https://orcid.org/0009-0006-1999-0317"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Mengyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102019411","display_name":"Haoyu Dong","orcid":"https://orcid.org/0009-0006-9789-7564","observed_orcids":["https://orcid.org/0009-0006-9789-7564","https://orcid.org/0000-0002-5132-0341"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dong, Haoyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102532315","display_name":"Shi Han","orcid":"https://orcid.org/0009-0006-0090-5944","observed_orcids":["https://orcid.org/0009-0006-0090-5944"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Shi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100331497","display_name":"Dongmei Zhang","orcid":"https://orcid.org/0000-0003-4560-9058","observed_orcids":["https://orcid.org/0000-0003-4560-9058"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Dongmei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T10028","display_name":"Topic Modeling","score":0.6948000192642212,"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":0.6948000192642212,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.08259999752044678,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.029899999499320984,"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/table-reasoning","display_name":"table reasoning","score":0.9649999737739563},{"id":"https://openalex.org/keywords/spreadsheet-formulas","display_name":"spreadsheet formulas","score":0.7020000219345093},{"id":"https://openalex.org/keywords/numerical-reasoning","display_name":"numerical reasoning","score":0.5759999752044678},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8029999732971191},{"id":"https://openalex.org/keywords/reward-signals","display_name":"reward signals","score":0.3779999911785126}],"concepts":[{"id":"https://openalex.org/C160145156","wikidata":"https://www.wikidata.org/wiki/Q778586","display_name":"Executable","level":2,"score":0.8492000102996826},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7142000198364258},{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.713100016117096},{"id":"https://openalex.org/C45235069","wikidata":"https://www.wikidata.org/wiki/Q278425","display_name":"Table (database)","level":2,"score":0.6675999760627747},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5557000041007996},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.42179998755455017},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.4120999872684479},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.3926999866962433},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.37630000710487366},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.37119999527931213},{"id":"https://openalex.org/C2776095079","wikidata":"https://www.wikidata.org/wiki/Q489538","display_name":"The Symbolic","level":2,"score":0.352400004863739},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.34929999709129333},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.30070000886917114},{"id":"https://openalex.org/C65620979","wikidata":"https://www.wikidata.org/wiki/Q7661176","display_name":"Symbolic data analysis","level":2,"score":0.2980000078678131},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.29760000109672546},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.289000004529953},{"id":"https://openalex.org/C2779639559","wikidata":"https://www.wikidata.org/wiki/Q7661178","display_name":"Symbolic execution","level":3,"score":0.27649998664855957},{"id":"https://openalex.org/C56949724","wikidata":"https://www.wikidata.org/wiki/Q219079","display_name":"Truth table","level":2,"score":0.27399998903274536},{"id":"https://openalex.org/C68476402","wikidata":"https://www.wikidata.org/wiki/Q1456936","display_name":"Table of contents","level":2,"score":0.27219998836517334}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2505.23667","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2505.23667","pdf_url":"https://arxiv.org/pdf/2505.23667","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.2505.23667","doi":"https://doi.org/10.48550/arxiv.2505.23667","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2505.23667","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:2505.23667","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2505.23667","pdf_url":"https://arxiv.org/pdf/2505.23667","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/W4416612957.pdf","grobid_xml":"https://content.openalex.org/works/W4416612957.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4295281117","https://openalex.org/W4411120194","https://openalex.org/W3199258251","https://openalex.org/W7128294665","https://openalex.org/W6966986356","https://openalex.org/W7117252075","https://openalex.org/W7168427491","https://openalex.org/W4415398054","https://openalex.org/W4414897670","https://openalex.org/W4406709267"],"abstract_inverted_index":{"Tables":[0],"are":[1],"a":[2,13,51,77,84],"fundamental":[3],"medium":[4],"for":[5,16,56,90,102,149],"organizing":[6],"and":[7,53,112,154],"analyzing":[8],"data,":[9,38,108],"making":[10],"table":[11,42,91,137,146,183],"reasoning":[12,27,35,62,138,194],"critical":[14],"capability":[15],"intelligent":[17],"systems.":[18],"Although":[19],"large":[20],"language":[21],"models":[22],"(LLMs)":[23],"exhibit":[24],"strong":[25],"general":[26,106],"abilities,":[28],"they":[29],"still":[30],"struggle":[31],"with":[32],"accurate":[33],"numerical":[34,156],"over":[36,105],"tabular":[37,107],"particularly":[39,148],"in":[40,181,196],"complex":[41,152],"settings":[43],"beyond":[44],"simple":[45],"relational":[46],"lookup.":[47],"Spreadsheet":[48],"formulas":[49,101],"provide":[50,174],"powerful":[52],"expressive":[54],"interface":[55],"executable":[57,99],"symbolic":[58],"operations,":[59],"enabling":[60],"rich":[61],"patterns":[63],"that":[64],"remain":[65],"largely":[66],"underexplored":[67],"by":[68],"existing":[69],"LLMs.":[70,197],"In":[71],"this":[72],"paper,":[73],"we":[74],"introduce":[75],"Formula-R1,":[76],"model":[78],"trained":[79],"via":[80],"Formula":[81,93,130],"Tuning":[82,94,131],"(Fortune),":[83],"formula-driven":[85,182,190],"reinforcement":[86],"learning":[87],"(RL)":[88],"framework":[89],"reasoning.":[92],"trains":[95],"LLMs":[96],"to":[97,192],"generate":[98],"spreadsheet":[100],"question":[103],"answering":[104],"using":[109],"execution":[110],"success":[111],"answer":[113],"correctness":[114],"as":[115],"reward":[116],"signals,":[117],"thereby":[118],"reducing":[119],"reliance":[120],"on":[121,135,145],"supervised":[122],"formula":[123],"annotations.":[124],"We":[125],"demonstrate":[126],"the":[127,177,186],"effectiveness":[128],"of":[129,179,189],"through":[132],"extensive":[133,172],"experiments":[134],"seven":[136],"benchmarks.":[139],"It":[140],"substantially":[141],"improves":[142],"LLM":[143],"performance":[144],"reasoning,":[147,184],"tasks":[150],"involving":[151],"tables":[153],"multi-step":[155],"computation.":[157],"Moreover,":[158],"Formula-R1":[159],"consistently":[160],"outperforms":[161],"prior":[162],"methods":[163],"under":[164],"controlled":[165],"comparison":[166],"settings.":[167],"Beyond":[168],"empirical":[169],"gains,":[170],"our":[171],"analyses":[173],"insights":[175],"into":[176],"role":[178],"RL":[180,191],"highlighting":[185],"broader":[187],"potential":[188],"enhance":[193],"capabilities":[195]},"counts_by_year":[],"updated_date":"2026-10-07T09:51:55.459714","created_date":"2025-10-10T00:00:00"}