{"id":"https://openalex.org/W7138348798","doi":"https://doi.org/10.1609/aaai.v40i37.40373","title":"HAPO: Training Language Models to Reason Concisely via History-Aware Policy Optimization","display_name":"HAPO: Training Language Models to Reason Concisely via History-Aware Policy Optimization","publication_year":2026,"publication_date":"2026-03-14","ids":{"openalex":"https://openalex.org/W7138348798","doi":"https://doi.org/10.1609/aaai.v40i37.40373"},"language":null,"primary_location":{"id":"doi:10.1609/aaai.v40i37.40373","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i37.40373","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/40373/44334","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://ojs.aaai.org/index.php/AAAI/article/download/40373/44334","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129646790","display_name":"Chengyu Huang","orcid":null},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chengyu Huang","raw_affiliation_strings":["Cornell University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cornell University","institution_ids":["https://openalex.org/I205783295"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129644835","display_name":"Zhengxin Zhang","orcid":null},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhengxin Zhang","raw_affiliation_strings":["Cornell University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cornell University","institution_ids":["https://openalex.org/I205783295"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5129684614","display_name":"Claire Cardie","orcid":null},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Claire Cardie","raw_affiliation_strings":["Cornell University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cornell University","institution_ids":["https://openalex.org/I205783295"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I205783295"],"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":"40","issue":"37","first_page":"31122","last_page":"31130"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.4837999939918518,"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":0.4837999939918518,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.14589999616146088,"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":0.060100000351667404,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/correctness","display_name":"Correctness","score":0.8230999708175659},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.7182999849319458},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4975000023841858},{"id":"https://openalex.org/keywords/state","display_name":"State (computer science)","score":0.46140000224113464},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4250999987125397},{"id":"https://openalex.org/keywords/function","display_name":"Function (biology)","score":0.4081999957561493}],"concepts":[{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.8230999708175659},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7253999710083008},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.7182999849319458},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5192000269889832},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4975000023841858},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.46140000224113464},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4250999987125397},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.4081999957561493},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.37439998984336853},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.32170000672340393},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.2953999936580658},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2815999984741211},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.2531000077724457},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25119999051094055}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1609/aaai.v40i37.40373","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i37.40373","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/40373/44334","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1609/aaai.v40i37.40373","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v40i37.40373","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/40373/44334","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7138348798.pdf","grobid_xml":"https://content.openalex.org/works/W7138348798.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"While":[0],"scaling":[1],"the":[2,14,58,95,120,144],"length":[3,47,97,110,156,202],"of":[4,19,90,122,146,204,209],"responses":[5,142],"at":[6],"test-time":[7,38],"has":[8],"been":[9],"shown":[10],"to":[11,70,118,172],"markedly":[12],"improve":[13],"reasoning":[15,199],"abilities":[16],"and":[17,30,167,176,178],"performance":[18],"large":[20],"language":[21],"models":[22],"(LLMs),":[23],"it":[24],"often":[25],"results":[26,191],"in":[27],"verbose":[28],"outputs":[29],"increases":[31],"inference":[32],"cost.":[33],"Prior":[34],"approaches":[35],"for":[36,103,165],"efficient":[37,151],"scaling,":[39],"typically":[40],"using":[41],"universal":[42],"budget":[43],"constraints":[44],"or":[45],"query-level":[46],"optimization,":[48],"do":[49],"not":[50],"leverage":[51],"historical":[52],"information":[53],"from":[54],"previous":[55],"encounters":[56],"with":[57,143,158,206],"same":[59],"problem":[60],"during":[61],"training.":[62],"We":[63,169],"hypothesize":[64],"that":[65,125,185,193],"this":[66,115,134,155],"limits":[67],"their":[68],"ability":[69],"progressively":[71],"make":[72],"solutions":[73,124],"more":[74,127,150],"concise":[75,128,198],"over":[76,98],"time.":[77],"To":[78],"address":[79],"this,":[80],"we":[81],"present":[82],"History-Aware":[83],"Policy":[84],"Optimization":[85],"(HAPO),":[86],"which":[87],"keeps":[88],"track":[89],"a":[91,108,159],"history":[92,116],"state":[93,117],"(e.g.,":[94],"minimum":[96],"previously":[99,131],"generated":[100],"correct":[101,123],"responses)":[102],"each":[104],"problem.":[105],"HAPO":[106,162,171,180,194],"employs":[107],"novel":[109],"reward":[111,135,157],"function":[112],"based":[113],"on":[114,181],"incentivize":[119],"discovery":[121],"are":[126],"than":[129],"those":[130],"found.":[132],"Crucially,":[133],"structure":[136],"avoids":[137],"overly":[138],"penalizing":[139],"shorter":[140],"incorrect":[141],"goal":[145],"facilitating":[147],"exploration":[148],"towards":[149],"solutions.":[152],"By":[153],"combining":[154],"correctness":[160,166],"reward,":[161],"jointly":[163],"optimizes":[164],"efficiency.":[168],"use":[170],"train":[173],"DeepSeek-R1-Distill-Qwen-1.5B,":[174],"DeepScaleR-1.5B-Preview,":[175],"Qwen-2.5-1.5B-Instruct,":[177],"evaluate":[179],"several":[182],"math":[183],"benchmarks":[184],"span":[186],"various":[187],"difficulty":[188],"levels.":[189],"Experiment":[190],"demonstrate":[192],"effectively":[195],"induces":[196],"LLMs\u2019":[197],"abilities,":[200],"producing":[201],"reductions":[203],"33-59%":[205],"accuracy":[207],"drops":[208],"only":[210],"2-5%.":[211]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-03-18T00:00:00"}