{"id":"https://openalex.org/W4390215524","doi":"https://doi.org/10.48550/arxiv.2312.14925","title":"A Survey of Reinforcement Learning from Human Feedback","display_name":"A Survey of Reinforcement Learning from Human Feedback","publication_year":2023,"publication_date":"2023-12-22","ids":{"openalex":"https://openalex.org/W4390215524","doi":"https://doi.org/10.48550/arxiv.2312.14925"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2312.14925","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2312.14925","pdf_url":"https://arxiv.org/pdf/2312.14925","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2312.14925","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5102866977","display_name":"Timo Kaufmann","orcid":"https://orcid.org/0000-0001-5193-8574","observed_orcids":["https://orcid.org/0000-0001-5193-8574"]},"institutions":[{"id":"https://openalex.org/I3018771216","display_name":"LMU Klinikum","ror":"https://ror.org/02jet3w32","country_code":"DE","type":"healthcare","lineage":["https://openalex.org/I3018771216","https://openalex.org/I8204097"]},{"id":"https://openalex.org/I8204097","display_name":"Ludwig-Maximilians-Universit\u00e4t M\u00fcnchen","ror":"https://ror.org/05591te55","country_code":"DE","type":"education","lineage":["https://openalex.org/I8204097"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Kaufmann, Timo","raw_affiliation_strings":["LMU Munich , MCML Munich"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LMU Munich , MCML Munich","institution_ids":["https://openalex.org/I3018771216","https://openalex.org/I8204097"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073106112","display_name":"Paul Weng","orcid":"https://orcid.org/0000-0002-2008-4569","observed_orcids":["https://orcid.org/0000-0002-2008-4569"]},"institutions":[{"id":"https://openalex.org/I3018771216","display_name":"LMU Klinikum","ror":"https://ror.org/02jet3w32","country_code":"DE","type":"healthcare","lineage":["https://openalex.org/I3018771216","https://openalex.org/I8204097"]},{"id":"https://openalex.org/I8204097","display_name":"Ludwig-Maximilians-Universit\u00e4t M\u00fcnchen","ror":"https://ror.org/05591te55","country_code":"DE","type":"education","lineage":["https://openalex.org/I8204097"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Weng, Paul","raw_affiliation_strings":["LMU Munich , MCML Munich"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LMU Munich , MCML Munich","institution_ids":["https://openalex.org/I3018771216","https://openalex.org/I8204097"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005338669","display_name":"Viktor Bengs","orcid":"https://orcid.org/0000-0001-6988-6186","observed_orcids":["https://orcid.org/0000-0001-6988-6186"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bengs, Viktor","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5059439673","display_name":"Eyke H\u00fcllermeier","orcid":"https://orcid.org/0000-0002-9944-4108","observed_orcids":["https://orcid.org/0000-0002-9944-4108"]},"institutions":[{"id":"https://openalex.org/I8204097","display_name":"Ludwig-Maximilians-Universit\u00e4t M\u00fcnchen","ror":"https://ror.org/05591te55","country_code":"DE","type":"education","lineage":["https://openalex.org/I8204097"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"H\u00fcllermeier, Eyke","raw_affiliation_strings":["LMU Munich , MCML Munich , DFKI Kaiserslautern"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LMU Munich , MCML Munich , DFKI Kaiserslautern","institution_ids":["https://openalex.org/I8204097"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":44,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9994999766349792,"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":0.9994999766349792,"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/T12072","display_name":"Machine Learning and Algorithms","score":9.999999747378752e-05,"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/T12101","display_name":"Advanced Bandit Algorithms Research","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/reinforcement-learning-from-human-feedback","display_name":"reinforcement learning from human feedback","score":0.9980000257492065},{"id":"https://openalex.org/keywords/preference-based-reinforcement-learning","display_name":"preference-based reinforcement learning","score":0.9879999756813049},{"id":"https://openalex.org/keywords/human-feedback","display_name":"human feedback","score":0.7139999866485596},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.9350000023841858},{"id":"https://openalex.org/keywords/human-computer-interaction","display_name":"human-computer interaction","score":0.6439999938011169},{"id":"https://openalex.org/keywords/ai-alignment","display_name":"AI alignment","score":0.6039999723434448},{"id":"https://openalex.org/keywords/human-values","display_name":"human values","score":0.609000027179718},{"id":"https://openalex.org/keywords/robot-control","display_name":"robot control","score":0.5730000138282776},{"id":"https://openalex.org/keywords/reward-function-learning","display_name":"reward function learning","score":0.4180000126361847},{"id":"https://openalex.org/keywords/human-in-the-loop-machine-learning","display_name":"human-in-the-loop learning","score":0.5889999866485596}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.8124520778656006},{"id":"https://openalex.org/C177606310","wikidata":"https://www.wikidata.org/wiki/Q5674297","display_name":"Adaptability","level":2,"score":0.5923762321472168},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.5625149607658386},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.5617844462394714},{"id":"https://openalex.org/C64543145","wikidata":"https://www.wikidata.org/wiki/Q162942","display_name":"Intersection (aeronautics)","level":2,"score":0.5511627793312073},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5417211651802063},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.5361376404762268},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.4428727328777313},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41353869438171387},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.3518891930580139},{"id":"https://openalex.org/C188147891","wikidata":"https://www.wikidata.org/wiki/Q147638","display_name":"Cognitive science","level":1,"score":0.34301263093948364},{"id":"https://openalex.org/C56739046","wikidata":"https://www.wikidata.org/wiki/Q192060","display_name":"Knowledge management","level":1,"score":0.3262108564376831},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.21938470005989075},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.2101033329963684},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.1300031840801239},{"id":"https://openalex.org/C78458016","wikidata":"https://www.wikidata.org/wiki/Q840400","display_name":"Evolutionary biology","level":1,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C146978453","wikidata":"https://www.wikidata.org/wiki/Q3798668","display_name":"Aerospace engineering","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2312.14925","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2312.14925","pdf_url":"https://arxiv.org/pdf/2312.14925","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2312.14925","doi":"https://doi.org/10.48550/arxiv.2312.14925","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2312.14925","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2312.14925","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2312.14925","pdf_url":"https://arxiv.org/pdf/2312.14925","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G2531449815","display_name":"Learning with Multiple Representations","funder_award_id":"101073307","funder_id":"https://openalex.org/F4320338438","funder_display_name":"HORIZON EUROPE Marie Sklodowska-Curie Actions"},{"id":"https://openalex.org/G3588374206","display_name":null,"funder_award_id":"62176154","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320315885","display_name":"Australian Government","ror":"https://ror.org/0314h5y94"},{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320321114","display_name":"Bundesministerium f\u00fcr Bildung und Forschung","ror":"https://ror.org/04pz7b180"},{"id":"https://openalex.org/F4320334322","display_name":"HORIZON EUROPE Framework Programme","ror":null},{"id":"https://openalex.org/F4320338438","display_name":"HORIZON EUROPE Marie Sklodowska-Curie Actions","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4390215524.pdf","grobid_xml":"https://content.openalex.org/works/W4390215524.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4408595355","https://openalex.org/W7150914366","https://openalex.org/W7171303336","https://openalex.org/W2944766483","https://openalex.org/W4385934332","https://openalex.org/W4387892197","https://openalex.org/W7038710371","https://openalex.org/W4411063470","https://openalex.org/W4414520407","https://openalex.org/W6950338373"],"abstract_inverted_index":{"Reinforcement":[0],"learning":[1,11,37],"from":[2,15],"human":[3,16,74,105,123,173],"feedback":[4,17,174],"(RLHF)":[5],"is":[6,187],"a":[7,53,96,158,193],"variant":[8],"of":[9,19,34,44,62,70,112,115,196],"reinforcement":[10,36],"(RL)":[12],"that":[13,167],"learns":[14],"instead":[18],"relying":[20],"on":[21,27,30,130],"an":[22,110],"engineered":[23],"reward":[24],"function.":[25],"Building":[26],"prior":[28],"work":[29,175],"the":[31,42,58,68,101,113,137,164,178,183],"related":[32],"setting":[33],"preference-based":[35],"(PbRL),":[38],"it":[39],"stands":[40],"at":[41],"intersection":[43],"artificial":[45],"intelligence":[46],"and":[47,60,150,172,177,191],"human-computer":[48],"interaction.":[49],"This":[50,107],"positioning":[51],"provides":[52,109],"promising":[54],"approach":[55],"to":[56,188],"enhance":[57],"performance":[59],"adaptability":[61],"intelligent":[63],"systems":[64],"while":[65],"also":[66],"improving":[67],"alignment":[69],"their":[71],"objectives":[72],"with":[73,122],"values.":[75],"The":[76],"success":[77],"in":[78,89,99,148,182],"training":[79],"large":[80],"language":[81],"models":[82],"(LLMs)":[83],"has":[84,94,128],"impressively":[85],"demonstrated":[86],"this":[87,197],"potential":[88],"recent":[90,126],"years,":[91],"where":[92,152],"RLHF":[93,131],"played":[95],"decisive":[97],"role":[98],"directing":[100],"model's":[102],"capabilities":[103],"towards":[104],"objectives.":[106],"article":[108],"overview":[111],"fundamentals":[114],"RLHF,":[116,169],"exploring":[117],"how":[118,170],"RL":[119],"agents":[120],"interact":[121],"feedback.":[124],"While":[125],"focus":[127],"been":[129],"for":[132],"LLMs,":[133],"our":[134,144],"survey":[135],"covers":[136],"technique":[138],"across":[139],"multiple":[140],"domains.":[141],"We":[142,162],"provide":[143],"most":[145],"comprehensive":[146],"coverage":[147],"control":[149],"robotics,":[151],"many":[153],"fundamental":[154],"techniques":[155],"originate,":[156],"alongside":[157],"dedicated":[159],"LLM":[160],"section.":[161],"examine":[163],"core":[165],"principles":[166],"underpin":[168],"algorithms":[171],"together,":[176],"main":[179],"research":[180],"trends":[181],"field.":[184,200],"Our":[185],"goal":[186],"give":[189],"researchers":[190],"practitioners":[192],"clear":[194],"understanding":[195],"rapidly":[198],"growing":[199]},"counts_by_year":[{"year":2026,"cited_by_count":4},{"year":2025,"cited_by_count":25},{"year":2024,"cited_by_count":15}],"updated_date":"2026-10-31T12:09:11.096373","created_date":"2023-12-26T00:00:00"}