{"id":"https://openalex.org/W7166811087","doi":"https://doi.org/10.18653/v1/2026.findings-acl.839","title":"Can Post-Training Transform LLMs into Causal Reasoners?","display_name":"Can Post-Training Transform LLMs into Causal Reasoners?","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166811087","doi":"https://doi.org/10.18653/v1/2026.findings-acl.839"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.839","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.839","pdf_url":"https://aclanthology.org/2026.findings-acl.839.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.839.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5103076254","display_name":"Junqi Chen","orcid":"https://orcid.org/0009-0008-0376-9356"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Junqi Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139753613","display_name":"Sirui Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sirui Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139793561","display_name":"Chaochao Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chaochao Lu","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.80120226,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"17020","last_page":"17038"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.21780000627040863,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.21780000627040863,"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/T11010","display_name":"Logic, Reasoning, and Knowledge","score":0.18799999356269836,"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/T13643","display_name":"Artificial Intelligence in Law","score":0.07400000095367432,"subfield":{"id":"https://openalex.org/subfields/3320","display_name":"Political Science and International Relations"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/causality","display_name":"Causality (physics)","score":0.29269999265670776},{"id":"https://openalex.org/keywords/product","display_name":"Product (mathematics)","score":0.25440001487731934},{"id":"https://openalex.org/keywords/ethnic-group","display_name":"Ethnic group","score":0.25279998779296875},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.24940000474452972},{"id":"https://openalex.org/keywords/population","display_name":"Population","score":0.2337000072002411}],"concepts":[{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.35929998755455017},{"id":"https://openalex.org/C64357122","wikidata":"https://www.wikidata.org/wiki/Q1149766","display_name":"Causality (physics)","level":2,"score":0.29269999265670776},{"id":"https://openalex.org/C90673727","wikidata":"https://www.wikidata.org/wiki/Q901718","display_name":"Product (mathematics)","level":2,"score":0.25440001487731934},{"id":"https://openalex.org/C137403100","wikidata":"https://www.wikidata.org/wiki/Q41710","display_name":"Ethnic group","level":2,"score":0.25279998779296875},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.24940000474452972},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.24469999969005585},{"id":"https://openalex.org/C2908647359","wikidata":"https://www.wikidata.org/wiki/Q2625603","display_name":"Population","level":2,"score":0.2337000072002411},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.2281000018119812},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.22470000386238098},{"id":"https://openalex.org/C2778137410","wikidata":"https://www.wikidata.org/wiki/Q2732820","display_name":"Government (linguistics)","level":2,"score":0.22139999270439148}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.839","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.839","pdf_url":"https://aclanthology.org/2026.findings-acl.839.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.839","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.839","pdf_url":"https://aclanthology.org/2026.findings-acl.839.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166811087.pdf","grobid_xml":"https://content.openalex.org/works/W7166811087.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Causal":[0],"inference":[1,102],"is":[2,36],"essential":[3],"for":[4,9,50,63],"decisionmaking":[5],"but":[6],"remains":[7],"challenging":[8],"nonexperts.While":[10],"large":[11],"language":[12],"models":[13],"(LLMs)":[14],"show":[15],"promise":[16],"in":[17],"this":[18,70],"domain,":[19],"their":[20],"precise":[21],"causal":[22,51,61,101,158],"estimation":[23],"capabilities":[24],"are":[25,163],"still":[26],"limited,":[27],"and":[28,65,82,86,130,139,155,161],"the":[29,41,115,124,145],"impact":[30],"of":[31],"post-training":[32,45,76,95,151],"on":[33,114],"these":[34,142],"abilities":[35],"insufficiently":[37],"explored.This":[38],"paper":[39],"examines":[40],"extent":[42],"to":[43,99,119],"which":[44],"can":[46,152],"enhance":[47],"LLMs'":[48],"capacity":[49],"inference.We":[52],"introduce":[53],"CauGym,":[54],"a":[55],"comprehensive":[56],"dataset":[57],"comprising":[58],"seven":[59],"core":[60],"tasks":[62],"training":[64],"five":[66,75,84],"diverse":[67],"test":[68],"sets.Using":[69],"dataset,":[71],"we":[72],"systematically":[73],"evaluate":[74],"approaches:":[77],"SFT,":[78],"DPO,":[79],"KTO,":[80],"PPO,":[81],"GRPO.Across":[83],"in-domain":[85],"four":[87],"existing":[88],"benchmarks,":[89],"our":[90],"experiments":[91],"demonstrate":[92],"that":[93,149],"appropriate":[94],"enables":[96],"smaller":[97],"LLMs":[98,126],"perform":[100],"competitively,":[103],"often":[104],"surpassing":[105],"much":[106],"larger":[107],"models.Our":[108],"14Bparameter":[109],"model":[110],"achieves":[111],"93.5%":[112],"accuracy":[113],"CaLM":[116],"benchmark,":[117],"compared":[118],"55.4%":[120],"by":[121],"OpenAI":[122],"o3.Furthermore,":[123],"post-trained":[125],"exhibit":[127],"strong":[128],"generalization":[129],"robustness":[131],"under":[132],"real-world":[133],"conditions":[134],"such":[135],"as":[136],"distribution":[137],"shifts":[138],"noisy":[140],"data.Collectively,":[141],"findings":[143],"provide":[144],"first":[146],"systematic":[147],"evidence":[148],"targeted":[150],"produce":[153],"reliable":[154],"robust":[156],"LLM-based":[157],"reasoners.Our":[159],"data":[160],"GRPO-model":[162],"available":[164],"at":[165]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}