{"id":"https://openalex.org/W4408356045","doi":"https://doi.org/10.1109/icassp49660.2025.10887891","title":"Online Learning from Strategic Human Feedback in LLM Fine-Tuning","display_name":"Online Learning from Strategic Human Feedback in LLM Fine-Tuning","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408356045","doi":"https://doi.org/10.1109/icassp49660.2025.10887891"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10887891","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10887891","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5028063502","display_name":"Shugang Hao","orcid":"https://orcid.org/0000-0003-2296-7004"},"institutions":[{"id":"https://openalex.org/I152815399","display_name":"Singapore University of Technology and Design","ror":"https://ror.org/05j6fvn87","country_code":"SG","type":"education","lineage":["https://openalex.org/I152815399"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Shugang Hao","raw_affiliation_strings":["Singapore University of Technology and Design,Pillar of Engineering Systems and Design"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Singapore University of Technology and Design,Pillar of Engineering Systems and Design","institution_ids":["https://openalex.org/I152815399"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5007119869","display_name":"Lingjie Duan","orcid":"https://orcid.org/0000-0002-0217-6507"},"institutions":[{"id":"https://openalex.org/I152815399","display_name":"Singapore University of Technology and Design","ror":"https://ror.org/05j6fvn87","country_code":"SG","type":"education","lineage":["https://openalex.org/I152815399"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Lingjie Duan","raw_affiliation_strings":["Singapore University of Technology and Design,Pillar of Engineering Systems and Design"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Singapore University of Technology and Design,Pillar of Engineering Systems and Design","institution_ids":["https://openalex.org/I152815399"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I152815399"],"apc_list":null,"apc_paid":null,"fwci":6.6921,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.96361341,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.24410000443458557,"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/T10320","display_name":"Neural Networks and Applications","score":0.24410000443458557,"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/computer-science","display_name":"Computer science","score":0.634617805480957},{"id":"https://openalex.org/keywords/online-learning","display_name":"Online learning","score":0.47810691595077515},{"id":"https://openalex.org/keywords/human\u2013computer-interaction","display_name":"Human\u2013computer interaction","score":0.4066484272480011},{"id":"https://openalex.org/keywords/multimedia","display_name":"Multimedia","score":0.2856988310813904}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.634617805480957},{"id":"https://openalex.org/C2986087404","wikidata":"https://www.wikidata.org/wiki/Q15946010","display_name":"Online learning","level":2,"score":0.47810691595077515},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.4066484272480011},{"id":"https://openalex.org/C49774154","wikidata":"https://www.wikidata.org/wiki/Q131765","display_name":"Multimedia","level":1,"score":0.2856988310813904}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10887891","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10887891","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":25,"referenced_works":["https://openalex.org/W2760781482","https://openalex.org/W3180245195","https://openalex.org/W4366341216","https://openalex.org/W4389520781","https://openalex.org/W4393159932","https://openalex.org/W4401857118","https://openalex.org/W6733923534","https://openalex.org/W6739585900","https://openalex.org/W6773817413","https://openalex.org/W6804385793","https://openalex.org/W6810738896","https://openalex.org/W6842525402","https://openalex.org/W6852418670","https://openalex.org/W6853300631","https://openalex.org/W6854866820","https://openalex.org/W6865090951","https://openalex.org/W6865844813","https://openalex.org/W6866459896","https://openalex.org/W6868737794","https://openalex.org/W6868805674","https://openalex.org/W6869412693","https://openalex.org/W6869444827","https://openalex.org/W6869781700","https://openalex.org/W6875735434","https://openalex.org/W7019882953"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Reinforcement":[0],"learning":[1,85],"from":[2],"human":[3,21,24,63,89,106],"feedback":[4,38,53,116],"(RLHF)":[5],"has":[6],"become":[7],"an":[8],"essential":[9],"step":[10],"in":[11,91,109],"fine-tuning":[12,94],"large":[13],"language":[14],"models":[15],"(LLMs)":[16],"to":[17,39,58,66,82],"align":[18],"them":[19],"with":[20],"preferences.":[22,31,47],"However,":[23],"labelers":[25,90],"are":[26,79],"selfish":[27],"and":[28,56,103,117],"have":[29],"diverse":[30],"They":[32],"may":[33],"strategically":[34],"misreport":[35],"their":[36,45,114],"online":[37,84],"influence":[40],"the":[41,60,80,92,110,129],"system\u2019s":[42],"aggregation":[43],"towards":[44],"own":[46],"Current":[48],"practice":[49],"simply":[50],"averages":[51],"labelers\u2019":[52,107],"per":[54],"time":[55,72],"fails":[57],"identify":[59],"most":[61],"accurate":[62],"labeler,":[64],"leading":[65],"linear":[67],"regret":[68,119],"${\\mathcal{O}}(T)$":[69],"for":[70],"T":[71],"slots.":[73],"To":[74],"our":[75,124],"best":[76],"knowledge,":[77],"we":[78],"first":[81],"study":[83],"mechanisms":[86],"against":[87],"strategic":[88],"LLM":[93],"process.":[95],"We":[96],"formulate":[97],"a":[98],"new":[99],"dynamic":[100],"Bayesian":[101],"game":[102],"dynamically":[104],"adjust":[105],"weights":[108],"preference":[111],"aggregation,":[112],"ensuring":[113],"truthful":[115],"sublinear":[118],"${\\mathcal{O}}\\left({{T^{1/2}}}\\right)$.":[120],"Simulation":[121],"results":[122],"demonstrate":[123],"mechanism\u2019s":[125],"great":[126],"advantages":[127],"over":[128],"existing":[130],"benchmark":[131],"schemes.":[132]},"counts_by_year":[{"year":2025,"cited_by_count":4}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}