{"id":"https://openalex.org/W4387724705","doi":"https://doi.org/10.48550/arxiv.2310.10076","title":"Verbosity Bias in Preference Labeling by Large Language Models","display_name":"Verbosity Bias in Preference Labeling by Large Language Models","publication_year":2023,"publication_date":"2023-10-16","ids":{"openalex":"https://openalex.org/W4387724705","doi":"https://doi.org/10.48550/arxiv.2310.10076"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2310.10076","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2310.10076","pdf_url":"https://arxiv.org/pdf/2310.10076","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":null,"license_id":null,"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/2310.10076","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5111058354","display_name":"K. Saito","orcid":"https://orcid.org/0000-0001-8910-9317","observed_orcids":["https://orcid.org/0000-0001-8910-9317"]},"institutions":[{"id":"https://openalex.org/I146399215","display_name":"University of Tsukuba","ror":"https://ror.org/02956yf07","country_code":"JP","type":"education","lineage":["https://openalex.org/I146399215"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Saito, Keita","raw_affiliation_strings":["University of Tsukuba & RIKEN AIP Tsukuba , Ibaraki 305-8573 , Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Tsukuba & RIKEN AIP Tsukuba , Ibaraki 305-8573 , Japan","institution_ids":["https://openalex.org/I146399215"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052503437","display_name":"Akifumi Wachi","orcid":"https://orcid.org/0009-0008-4722-9861","observed_orcids":["https://orcid.org/0009-0008-4722-9861"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wachi, Akifumi","raw_affiliation_strings":["LY Corporation Chiyoda-ku , Tokyo 102-8282 , Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LY Corporation Chiyoda-ku , Tokyo 102-8282 , Japan","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114285442","display_name":"Koki Wataoka","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wataoka, Koki","raw_affiliation_strings":["LY Corporation Chiyoda-ku , Tokyo 102-8282 , Japan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"LY Corporation Chiyoda-ku , Tokyo 102-8282 , Japan","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5038757231","display_name":"Youhei Akimoto","orcid":"https://orcid.org/0000-0003-2760-8123","observed_orcids":["https://orcid.org/0000-0003-2760-8123"]},"institutions":[{"id":"https://openalex.org/I146399215","display_name":"University of Tsukuba","ror":"https://ror.org/02956yf07","country_code":"JP","type":"education","lineage":["https://openalex.org/I146399215"]}],"countries":["JP"],"is_corresponding":false,"raw_author_name":"Akimoto, Youhei","raw_affiliation_strings":["RIKEN AIP Tsukuba , Ibaraki 305-8573 , Japan","University of Tsukuba &"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"RIKEN AIP Tsukuba , Ibaraki 305-8573 , Japan","institution_ids":[]},{"raw_affiliation_string":"University of Tsukuba &","institution_ids":["https://openalex.org/I146399215"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":14,"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/T10028","display_name":"Topic Modeling","score":0.37790000438690186,"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/T10028","display_name":"Topic Modeling","score":0.37790000438690186,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.3495999872684479,"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/T10883","display_name":"Ethics and Social Impacts of AI","score":0.10140000283718109,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.9549999833106995},{"id":"https://openalex.org/keywords/reinforcement-learning-from-ai-feedback","display_name":"reinforcement learning from AI feedback","score":0.8349999785423279},{"id":"https://openalex.org/keywords/reinforcement-learning-from-human-feedback","display_name":"reinforcement learning from human feedback","score":0.621999979019165},{"id":"https://openalex.org/keywords/gpt-4","display_name":"GPT-4","score":0.7559999823570251},{"id":"https://openalex.org/keywords/bias-metrics","display_name":"bias metrics","score":0.5450000166893005},{"id":"https://openalex.org/keywords/answer-length","display_name":"answer length","score":0.5640000104904175}],"concepts":[{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.5615212321281433},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.4753146767616272},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.4364740252494812},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.39963704347610474},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.35004985332489014},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.2337377965450287},{"id":"https://openalex.org/C175444787","wikidata":"https://www.wikidata.org/wiki/Q39072","display_name":"Microeconomics","level":1,"score":0.10291016101837158},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.1018511950969696}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2310.10076","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2310.10076","pdf_url":"https://arxiv.org/pdf/2310.10076","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2310.10076","doi":"https://doi.org/10.48550/arxiv.2310.10076","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2310.10076","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2310.10076","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2310.10076","pdf_url":"https://arxiv.org/pdf/2310.10076","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality education","score":0.8100000023841858}],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4387724705.pdf","grobid_xml":"https://content.openalex.org/works/W4387724705.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2748952813","https://openalex.org/W4391375266","https://openalex.org/W2931662336","https://openalex.org/W2077865380","https://openalex.org/W2765597752","https://openalex.org/W2134894512","https://openalex.org/W2083375246","https://openalex.org/W2090624569","https://openalex.org/W2067108088","https://openalex.org/W2010073985"],"abstract_inverted_index":{"In":[0,54],"recent":[1,56],"years,":[2],"Large":[3],"Language":[4],"Models":[5],"(LLMs)":[6],"have":[7,111],"witnessed":[8],"a":[9,92,99,131],"remarkable":[10],"surge":[11],"in":[12,27,117],"prevalence,":[13],"altering":[14],"the":[15,29,60,79],"landscape":[16],"of":[17,31,62],"natural":[18],"language":[19],"processing":[20],"and":[21,90],"machine":[22],"learning.":[23],"One":[24],"key":[25],"factor":[26],"improving":[28],"performance":[30],"LLMs":[32,48,69,86,89,102],"is":[33],"alignment":[34],"with":[35,38,65,84,87],"humans":[36],"achieved":[37],"Reinforcement":[39,71],"Learning":[40,72],"from":[41,67,73],"Human":[42],"Feedback":[43,75],"(RLHF),":[44],"as":[45,50],"for":[46],"many":[47],"such":[49],"GPT-4,":[51],"Bard,":[52],"etc.":[53],"addition,":[55],"studies":[57],"are":[58],"investigating":[59],"replacement":[61],"human":[63],"feedback":[64,66],"other":[68,88],"named":[70],"AI":[74],"(RLAIF).":[76],"We":[77,114,128],"examine":[78],"biases":[80],"that":[81,116],"come":[82],"along":[83],"evaluating":[85],"take":[91],"closer":[93],"look":[94],"into":[95],"verbosity":[96],"bias":[97,100],"--":[98],"where":[101],"sometimes":[103],"prefer":[104],"more":[105,125],"verbose":[106],"answers":[107,124],"even":[108],"if":[109],"they":[110],"similar":[112],"qualities.":[113],"see":[115],"our":[118],"problem":[119],"setting,":[120],"GPT-4":[121],"prefers":[122],"longer":[123],"than":[126],"humans.":[127],"also":[129],"propose":[130],"metric":[132],"to":[133],"measure":[134],"this":[135],"bias.":[136]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":7},{"year":2024,"cited_by_count":5}],"updated_date":"2026-10-16T09:59:22.235173","created_date":"2025-10-10T00:00:00"}