{"id":"https://openalex.org/W7166877047","doi":"https://doi.org/10.18653/v1/2026.acl-long.2206","title":"Influence-based Online Experience Selection for Effective RLHF","display_name":"Influence-based Online Experience Selection for Effective RLHF","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166877047","doi":"https://doi.org/10.18653/v1/2026.acl-long.2206"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.acl-long.2206","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.2206","pdf_url":"https://aclanthology.org/2026.acl-long.2206.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.2206.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139796889","display_name":"Yifan Gong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yifan Gong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139796887","display_name":"Jing Yao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jing Yao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139791005","display_name":"Xiting Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiting Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139784778","display_name":"Xunlong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xunlong Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026595284","display_name":"Xiaoyuan Yi","orcid":"https://orcid.org/0000-0003-2710-1613"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiaoyuan Yi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139740186","display_name":"Xing Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xing Xie","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":0.91723511,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"47755","last_page":"47771"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13353","display_name":"Flow Experience in Various Fields","score":0.12759999930858612,"subfield":{"id":"https://openalex.org/subfields/3204","display_name":"Developmental and Educational Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T13353","display_name":"Flow Experience in Various Fields","score":0.12759999930858612,"subfield":{"id":"https://openalex.org/subfields/3204","display_name":"Developmental and Educational 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/T10203","display_name":"Recommender Systems and Techniques","score":0.06639999896287918,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10803","display_name":"Innovative Human-Technology Interaction","score":0.06620000302791595,"subfield":{"id":"https://openalex.org/subfields/1709","display_name":"Human-Computer Interaction"},"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/selection","display_name":"Selection (genetic algorithm)","score":0.5235000252723694},{"id":"https://openalex.org/keywords/feature-selection","display_name":"Feature selection","score":0.2793999910354614},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.2754000127315521},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.24779999256134033},{"id":"https://openalex.org/keywords/identification","display_name":"Identification (biology)","score":0.24120000004768372}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5842000246047974},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.5235000252723694},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41019999980926514},{"id":"https://openalex.org/C148483581","wikidata":"https://www.wikidata.org/wiki/Q446488","display_name":"Feature selection","level":2,"score":0.2793999910354614},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2754000127315521},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.24779999256134033},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.24480000138282776},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.24140000343322754},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.24120000004768372},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.24009999632835388}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.2206","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.2206","pdf_url":"https://aclanthology.org/2026.acl-long.2206.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.2206","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.2206","pdf_url":"https://aclanthology.org/2026.acl-long.2206.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/G3248486410","display_name":null,"funder_award_id":"U2436209","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G4128466783","display_name":null,"funder_award_id":"92470205","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G6750762792","display_name":null,"funder_award_id":"62476279","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/F4320322499","display_name":"Renmin University of China","ror":"https://ror.org/041pakw92"},{"id":"https://openalex.org/F4320335787","display_name":"Fundamental Research Funds for the Central Universities","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166877047.pdf","grobid_xml":"https://content.openalex.org/works/W7166877047.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Reinforcement":[0],"Learning":[1],"from":[2],"Human":[3],"Feedback":[4],"(RLHF)":[5],"has":[6],"emerged":[7],"as":[8],"a":[9,92],"crucial":[10],"technique":[11],"for":[12,55,65,97],"aligning":[13],"large":[14],"language":[15],"models":[16],"(LLMs)":[17],"with":[18,138,158],"human":[19],"preferences.However,":[20],"existing":[21],"RLHF":[22,98],"methods":[23,64],"face":[24],"key":[25],"challenges,":[26],"including":[27],"poor":[28],"sample":[29],"efficiency,":[30],"high":[31],"computational":[32],"overhead,":[33],"and":[34,79,121],"slow":[35],"convergence.Recent":[36],"studies":[37],"highlight":[38],"the":[39,51,102,115,127],"importance":[40],"of":[41,104],"data":[42,62,78,94,112],"selection":[43,63,95],"in":[44],"RL,":[45],"but":[46],"how":[47],"to":[48,72],"effectively":[49],"select":[50],"most":[52],"beneficial":[53],"experiences":[54],"RL":[56,66,140],"training":[57,106,152],"remains":[58],"an":[59,74],"open":[60],"problem.Existing":[61],"rely":[67],"on":[68,108,126],"heuristic":[69],"metrics,":[70],"failing":[71],"establish":[73],"interpretable":[75],"connection":[76],"between":[77],"optimization":[80,160],"objectives.To":[81],"address":[82],"this":[83],"problem,":[84],"we":[85],"propose":[86],"InfOES":[87,118,149],"(Influence-based":[88],"Online":[89],"Experience":[90],"Selection),":[91],"novel":[93],"method":[96],"that":[99,148],"dynamically":[100],"estimates":[101],"influence":[103],"individual":[105],"samples":[107,125],"policy":[109,116],"optimization.By":[110],"incorporating":[111],"attribution":[113],"into":[114],"gradient,":[117],"can":[119],"identify":[120],"filter":[122],"out":[123],"detrimental":[124],"fly,":[128],"ensuring":[129],"effective":[130],"convergence":[131],"toward":[132],"alignment":[133,156],"objectives.Our":[134],"approach":[135],"is":[136],"compatible":[137],"various":[139],"algorithms":[141],"(e.g.,":[142],"PPO,":[143],"GRPO,":[144],"RE-INFORCE++).Extensive":[145],"experiments":[146],"demonstrate":[147],"significantly":[150],"enhances":[151],"effectiveness,":[153],"achieving":[154],"superior":[155],"performance":[157],"fewer":[159],"steps.":[161]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}