{"id":"https://openalex.org/W7166811811","doi":"https://doi.org/10.18653/v1/2026.acl-long.847","title":"Team-Based Self-Play With Dual Adaptive Weighting for Fine-Tuning LLMs","display_name":"Team-Based Self-Play With Dual Adaptive Weighting for Fine-Tuning LLMs","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166811811","doi":"https://doi.org/10.18653/v1/2026.acl-long.847"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.acl-long.847","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.847","pdf_url":"https://aclanthology.org/2026.acl-long.847.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":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.acl-long.847.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139816213","display_name":"Wu Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu Li","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123973721","display_name":"Yigeng Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yigeng Zhou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139823577","display_name":"Zesheng Shi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zesheng Shi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029640741","display_name":"Yequan Wang","orcid":"https://orcid.org/0000-0001-7530-6125"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yequan Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139762730","display_name":"Min Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Min Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139845544","display_name":"Jing Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jing Li","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":0.89812067,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"18617","last_page":"18632"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12784","display_name":"Modular Robots and Swarm Intelligence","score":0.2964000105857849,"subfield":{"id":"https://openalex.org/subfields/2210","display_name":"Mechanical Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12784","display_name":"Modular Robots and Swarm Intelligence","score":0.2964000105857849,"subfield":{"id":"https://openalex.org/subfields/2210","display_name":"Mechanical Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10970","display_name":"Team Dynamics and Performance","score":0.11500000208616257,"subfield":{"id":"https://openalex.org/subfields/3207","display_name":"Social Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10653","display_name":"Robot Manipulation and Learning","score":0.0421999990940094,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/weighting","display_name":"Weighting","score":0.4406999945640564},{"id":"https://openalex.org/keywords/dual","display_name":"Dual (grammatical number)","score":0.3946000039577484},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.3050000071525574},{"id":"https://openalex.org/keywords/control-theory","display_name":"Control theory (sociology)","score":0.26980000734329224}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.46369999647140503},{"id":"https://openalex.org/C183115368","wikidata":"https://www.wikidata.org/wiki/Q856577","display_name":"Weighting","level":2,"score":0.4406999945640564},{"id":"https://openalex.org/C2780980858","wikidata":"https://www.wikidata.org/wiki/Q110022","display_name":"Dual (grammatical number)","level":2,"score":0.3946000039577484},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3208000063896179},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.3050000071525574},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.2782999873161316},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.27619999647140503},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2732999920845032},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.26980000734329224},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.249099999666214}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.847","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.847","pdf_url":"https://aclanthology.org/2026.acl-long.847.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":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.847","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.847","pdf_url":"https://aclanthology.org/2026.acl-long.847.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":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G5516094551","display_name":null,"funder_award_id":"62476070","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320336618","display_name":"Science and Technology Program of Suzhou","ror":null}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166811811.pdf","grobid_xml":"https://content.openalex.org/works/W7166811811.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"While":[0],"recent":[1],"self-training":[2],"approaches":[3],"have":[4],"reduced":[5],"reliance":[6],"on":[7],"human-labeled":[8],"data":[9,22],"for":[10],"aligning":[11],"LLMs,":[12],"they":[13],"still":[14],"face":[15],"critical":[16],"limitations:":[17],"(i)":[18,107],"sensitivity":[19],"to":[20,25,37,66],"synthetic":[21],"quality,":[23],"leading":[24],"instability":[26],"and":[27,43,87,95,119,159],"bias":[28],"amplification":[29],"in":[30,69,78],"iterative":[31],"training;":[32],"(ii)":[33,120],"ineffective":[34],"optimization":[35],"due":[36],"a":[38,61,70,75,108,121,135],"diminishing":[39],"gap":[40],"between":[41],"positive":[42],"negative":[44],"responses":[45],"over":[46],"successive":[47],"training":[48],"iterations.In":[49],"this":[50],"paper,":[51],"we":[52,101],"propose":[53],"Team-based":[54],"self-Play":[55],"with":[56,86],"dual":[57],"Adaptive":[58],"Weighting":[59],"(TPAW),":[60],"novel":[62],"self-play":[63],"algorithm":[64],"designed":[65],"improve":[67],"alignment":[68,141],"fully":[71],"self-supervised":[72],"setting.TPAW":[73],"adopts":[74],"team-based":[76],"framework":[77],"which":[79],"the":[80,114],"current":[81],"policy":[82],"model":[83],"both":[84],"collaborates":[85],"competes":[88],"against":[89],"historical":[90],"checkpoints,":[91],"promoting":[92],"more":[93],"stable":[94],"efficient":[96],"optimization.To":[97],"further":[98],"enhance":[99],"learning,":[100],"design":[102],"two":[103],"adaptive":[104],"weighting":[105,123],"mechanisms:":[106],"response":[109],"reweighting":[110],"scheme":[111],"that":[112,125,149],"adjusts":[113],"importance":[115],"of":[116],"target":[117],"responses,":[118],"player":[122],"strategy":[124],"dynamically":[126],"modulates":[127],"each":[128],"team":[129],"member's":[130],"contribution":[131],"during":[132],"training.Initialized":[133],"from":[134],"SFT":[136],"model,":[137],"TPAW":[138,150],"iteratively":[139],"refines":[140],"without":[142],"requiring":[143],"additional":[144],"human":[145],"supervision.Experimental":[146],"results":[147],"demonstrate":[148],"consistently":[151],"outperforms":[152],"existing":[153],"baselines":[154],"across":[155],"various":[156],"base":[157],"models":[158],"LLM":[160],"benchmarks.":[161]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}