{"id":"https://openalex.org/W4413157667","doi":"https://doi.org/10.1109/tit.2025.3598296","title":"Sample-Efficient Reinforcement Learning From Human Feedback via Information-Directed Sampling","display_name":"Sample-Efficient Reinforcement Learning From Human Feedback via Information-Directed Sampling","publication_year":2025,"publication_date":"2025-08-13","ids":{"openalex":"https://openalex.org/W4413157667","doi":"https://doi.org/10.1109/tit.2025.3598296"},"language":"en","primary_location":{"id":"doi:10.1109/tit.2025.3598296","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tit.2025.3598296","pdf_url":null,"source":{"id":"https://openalex.org/S4502562","display_name":"IEEE Transactions on Information Theory","issn_l":"0018-9448","issn":["0018-9448","1557-9654"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Information Theory","raw_type":"journal-article"},"type":"article","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":null,"display_name":"Han Qi","orcid":"https://orcid.org/0009-0001-7684-9882"},"institutions":[{"id":"https://openalex.org/I87445476","display_name":"Xi'an Jiaotong University","ror":"https://ror.org/017zhmm22","country_code":"CN","type":"education","lineage":["https://openalex.org/I87445476"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Han Qi","raw_affiliation_strings":["School of Software Engineering, Xi&#x2019;an Jiaotong University, Xi&#x2019;an, China","School of Software Engineering, Xi&#x2019;an Jiaotong University, China"],"raw_orcid":"https://orcid.org/0009-0001-7684-9882","affiliations":[{"raw_affiliation_string":"School of Software Engineering, Xi&#x2019;an Jiaotong University, Xi&#x2019;an, China","institution_ids":["https://openalex.org/I87445476"]},{"raw_affiliation_string":"School of Software Engineering, Xi&#x2019;an Jiaotong University, China","institution_ids":["https://openalex.org/I87445476"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019772747","display_name":"Haochen Yang","orcid":null},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Haochen Yang","raw_affiliation_strings":["School of Mathematics, Peking University, Beijing, China","School of Mathematics, Peking University, China"],"raw_orcid":"https://orcid.org/0009-0009-1772-4172","affiliations":[{"raw_affiliation_string":"School of Mathematics, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Mathematics, Peking University, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081325393","display_name":"Qiaosheng Zhang","orcid":"https://orcid.org/0000-0001-6114-8453"},"institutions":[{"id":"https://openalex.org/I4210100255","display_name":"Beijing Academy of Artificial Intelligence","ror":"https://ror.org/016a74861","country_code":"CN","type":"other","lineage":["https://openalex.org/I4210100255"]},{"id":"https://openalex.org/I4391012619","display_name":"Shanghai Artificial Intelligence Laboratory","ror":"https://ror.org/03wkvpx79","country_code":null,"type":"facility","lineage":["https://openalex.org/I4391012619"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qiaosheng Zhang","raw_affiliation_strings":["Shanghai Artificial Intelligence Laboratory, Shanghai, China","Shanghai Artificial Intelligence Laboratory, China"],"raw_orcid":"https://orcid.org/0000-0001-6114-8453","affiliations":[{"raw_affiliation_string":"Shanghai Artificial Intelligence Laboratory, Shanghai, China","institution_ids":["https://openalex.org/I4210100255","https://openalex.org/I4391012619"]},{"raw_affiliation_string":"Shanghai Artificial Intelligence Laboratory, China","institution_ids":["https://openalex.org/I4210100255","https://openalex.org/I4391012619"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101727948","display_name":"Zhuoran Yang","orcid":"https://orcid.org/0000-0001-5269-9958"},"institutions":[{"id":"https://openalex.org/I32971472","display_name":"Yale University","ror":"https://ror.org/03v76x132","country_code":"US","type":"education","lineage":["https://openalex.org/I32971472"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhuoran Yang","raw_affiliation_strings":["Department of Statistics and Data Science, Yale University, New Haven, CT, USA","Department of Statistics and Data Science, Yale University, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Statistics and Data Science, Yale University, New Haven, CT, USA","institution_ids":["https://openalex.org/I32971472"]},{"raw_affiliation_string":"Department of Statistics and Data Science, Yale University, China","institution_ids":["https://openalex.org/I32971472"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I87445476"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.1006682,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"71","issue":"10","first_page":"7942","last_page":"7958"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.8870000243186951,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.8870000243186951,"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/T12879","display_name":"Distributed Sensor Networks and Detection Algorithms","score":0.8145999908447266,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T10320","display_name":"Neural Networks and Applications","score":0.7989000082015991,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.6770200133323669},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6438020467758179},{"id":"https://openalex.org/keywords/sample","display_name":"Sample (material)","score":0.5271536707878113},{"id":"https://openalex.org/keywords/sampling","display_name":"Sampling (signal processing)","score":0.5208052396774292},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4398185610771179},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.3563670516014099},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.35405874252319336},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.2807677388191223},{"id":"https://openalex.org/keywords/computer-vision","display_name":"Computer vision","score":0.11515593528747559}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6770200133323669},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6438020467758179},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.5271536707878113},{"id":"https://openalex.org/C140779682","wikidata":"https://www.wikidata.org/wiki/Q210868","display_name":"Sampling (signal processing)","level":3,"score":0.5208052396774292},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4398185610771179},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3563670516014099},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.35405874252319336},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2807677388191223},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.11515593528747559},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.0},{"id":"https://openalex.org/C43617362","wikidata":"https://www.wikidata.org/wiki/Q170050","display_name":"Chromatography","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tit.2025.3598296","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tit.2025.3598296","pdf_url":null,"source":{"id":"https://openalex.org/S4502562","display_name":"IEEE Transactions on Information Theory","issn_l":"0018-9448","issn":["0018-9448","1557-9654"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Information Theory","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Reduced inequalities","id":"https://metadata.un.org/sdg/10","score":0.4300000071525574}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4306904969","https://openalex.org/W2138720691","https://openalex.org/W4362501864","https://openalex.org/W4380318855","https://openalex.org/W3084456289","https://openalex.org/W2024136090","https://openalex.org/W4391331176","https://openalex.org/W2031695474","https://openalex.org/W2042102171","https://openalex.org/W3107697994"],"abstract_inverted_index":{"We":[0],"study":[1],"the":[2,27,50,53,65,70,74,83,108,145,152,166,170,174,200,220,242,250,260],"problem":[3,13],"of":[4,29,52,64,85,129,154,169,181,202,227,252,262],"reinforcement":[5,256],"learning":[6,257],"from":[7,19],"human":[8,78],"feedback":[9,79],"(RLHF),":[10],"a":[11,20,57,95,103,124],"critical":[12],"in":[14,235,255,259],"training":[15,261],"large":[16,86,263],"language":[17,264],"models,":[18],"theoretical":[21],"perspective.":[22],"Our":[23,47],"main":[24],"contribution":[25],"is":[26,144,151,163,180,213,231],"design":[28,225],"novel":[30,104],"sample-efficient":[31],"RLHF":[32,236],"algorithms":[33,48],"based":[34],"on":[35],"information-directed":[36],"sampling":[37],"(IDS),":[38],"an":[39,209],"online":[40],"decision-making":[41],"principle":[42,226],"inspired":[43],"by":[44],"information":[45,59,71,253],"theory.":[46],"maximize":[49],"sum":[51],"value":[54,251],"function":[55],"and":[56,89,101,156,195,204,258],"mutual":[58],"term":[60],"that":[61,212],"encourages":[62],"exploration":[63],"unknown":[66],"environment":[67,75],"(which":[68],"quantifies":[69],"gained":[72],"about":[73],"through":[76],"observed":[77],"data).":[80],"To":[81],"tackle":[82],"challenge":[84],"state":[87],"spaces":[88],"improve":[90],"sample":[91,222],"efficiency,":[92],"we":[93,207],"construct":[94],"simplified":[96],"<italic":[97,141,148,192,196],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[98,110,116,132,142,149,158,184,193,197],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">surrogate":[99],"environment</i>":[100],"introduce":[102],"distance":[105],"measure":[106],"(named":[107],"<inline-formula":[109,131,157,183],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">":[111,133,159,185],"<tex-math":[112,134,160,186],"notation=\"LaTeX\">$\\ell":[113],"_{g}$":[114],"</tex-math></inline-formula><italic":[115],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">-distance</i>),":[117],"enabling":[118],"our":[119,247],"IDS-based":[120],"algorithm":[121,211,230],"to":[122,165,173,241],"achieve":[123],"Bayesian":[125],"regret":[126,178],"upper":[127],"bound":[128,179],"order":[130,182],"notation=\"LaTeX\">$O(H^{3/2}\\sqrt":[135],"{\\log":[136],"(K(\\epsilon))":[137],"T})$":[138],"</tex-math></inline-formula>,":[139,190],"where":[140,191],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">H</i>":[143],"episode":[146,155],"length,":[147],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">T</i>":[150],"number":[153,168],"notation=\"LaTeX\">$K(\\epsilon)$":[161],"</tex-math></inline-formula>":[162],"related":[164],"covering":[167],"environment.":[171],"Specializing":[172],"tabular":[175],"settings,":[176],"this":[177,228],"notation=\"LaTeX\">$\\tilde":[187],"{O}(H^{2}\\sqrt":[188],"{SAT})$":[189],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">S</i>":[194],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">A</i>":[198],"are":[199],"numbers":[201],"states":[203],"actions.":[205],"Finally,":[206],"propose":[208],"Approximate-IDS":[210],"computationally":[214],"more":[215],"efficient":[216],"while":[217],"maintaining":[218],"nearly":[219],"same":[221],"efficiency.":[223],"The":[224],"approximate":[229],"not":[232],"only":[233],"effective":[234],"settings":[237],"but":[238],"also":[239],"applicable":[240],"standard":[243],"RL":[244],"framework.":[245],"Moreover,":[246],"work":[248],"showcases":[249],"theory":[254],"models.":[265]},"counts_by_year":[],"updated_date":"2025-12-21T23:12:01.093139","created_date":"2025-10-10T00:00:00"}