{"id":"https://openalex.org/W7133530826","doi":"https://doi.org/10.48550/arxiv.2603.02232","title":"Beyond Binary Preferences: A Principled Framework for Reward Modeling with Ordinal Feedback","display_name":"Beyond Binary Preferences: A Principled Framework for Reward Modeling with Ordinal Feedback","publication_year":2026,"publication_date":"2026-02-13","ids":{"openalex":"https://openalex.org/W7133530826","doi":"https://doi.org/10.48550/arxiv.2603.02232"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.02232","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02232","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.02232","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5128065176","display_name":"Amirhossein Afsharrad","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Afsharrad, Amirhossein","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029262439","display_name":"Ruida Zhou","orcid":"https://orcid.org/0000-0002-8855-2031"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Ruida","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127837741","display_name":"Luca Viano","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Viano, Luca","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112601460","display_name":"S. Lall","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lall, Sanjay","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5128112337","display_name":"Mohammad Ghavamzadeh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ghavamzadeh, Mohammad","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":null,"has_fulltext":false,"cited_by_count":0,"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/T10667","display_name":"Emotion and Mood Recognition","score":0.10760000348091125,"subfield":{"id":"https://openalex.org/subfields/3205","display_name":"Experimental and Cognitive 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/T10667","display_name":"Emotion and Mood Recognition","score":0.10760000348091125,"subfield":{"id":"https://openalex.org/subfields/3205","display_name":"Experimental and Cognitive 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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.08320000022649765,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.060600001364946365,"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/ordinal-regression","display_name":"Ordinal regression","score":0.7444000244140625},{"id":"https://openalex.org/keywords/ordinal-data","display_name":"Ordinal data","score":0.6624000072479248},{"id":"https://openalex.org/keywords/probabilistic-logic","display_name":"Probabilistic logic","score":0.6093999743461609},{"id":"https://openalex.org/keywords/heuristic","display_name":"Heuristic","score":0.582099974155426},{"id":"https://openalex.org/keywords/preference","display_name":"Preference","score":0.5368000268936157},{"id":"https://openalex.org/keywords/likert-scale","display_name":"Likert scale","score":0.48590001463890076},{"id":"https://openalex.org/keywords/binary-number","display_name":"Binary number","score":0.48500001430511475},{"id":"https://openalex.org/keywords/margin","display_name":"Margin (machine learning)","score":0.4641000032424927}],"concepts":[{"id":"https://openalex.org/C110313322","wikidata":"https://www.wikidata.org/wiki/Q7100793","display_name":"Ordinal regression","level":2,"score":0.7444000244140625},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6802999973297119},{"id":"https://openalex.org/C85461838","wikidata":"https://www.wikidata.org/wiki/Q7100785","display_name":"Ordinal data","level":2,"score":0.6624000072479248},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.6093999743461609},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.582099974155426},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.5368000268936157},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4968000054359436},{"id":"https://openalex.org/C105776082","wikidata":"https://www.wikidata.org/wiki/Q617473","display_name":"Likert scale","level":2,"score":0.48590001463890076},{"id":"https://openalex.org/C48372109","wikidata":"https://www.wikidata.org/wiki/Q3913","display_name":"Binary number","level":2,"score":0.48500001430511475},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.4641000032424927},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.44040000438690186},{"id":"https://openalex.org/C2909711754","wikidata":"https://www.wikidata.org/wiki/Q7100785","display_name":"Ordinal Scale","level":2,"score":0.3720000088214874},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.34540000557899475},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.33070001006126404},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.3264000117778778},{"id":"https://openalex.org/C91682802","wikidata":"https://www.wikidata.org/wiki/Q620538","display_name":"Multidimensional scaling","level":2,"score":0.3244999945163727},{"id":"https://openalex.org/C152877465","wikidata":"https://www.wikidata.org/wiki/Q208042","display_name":"Regression analysis","level":2,"score":0.3203999996185303},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.30160000920295715},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.29120001196861267},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.28780001401901245},{"id":"https://openalex.org/C83546350","wikidata":"https://www.wikidata.org/wiki/Q1139051","display_name":"Regression","level":2,"score":0.27720001339912415},{"id":"https://openalex.org/C2777868144","wikidata":"https://www.wikidata.org/wiki/Q7239817","display_name":"Preference elicitation","level":3,"score":0.2687999904155731},{"id":"https://openalex.org/C2779190172","wikidata":"https://www.wikidata.org/wiki/Q4913888","display_name":"Binary data","level":3,"score":0.2590000033378601},{"id":"https://openalex.org/C76969082","wikidata":"https://www.wikidata.org/wiki/Q486902","display_name":"Mathematical model","level":2,"score":0.2542000114917755},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.25200000405311584}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.02232","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02232","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.02232","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02232","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,"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Reward":[0],"modeling":[1,89],"is":[2,78],"crucial":[3],"for":[4,20,73,195],"aligning":[5],"large":[6],"language":[7],"models":[8,63,211],"with":[9,90],"human":[10,26,219],"preferences,":[11],"yet":[12],"current":[13],"approaches":[14,67],"lack":[15,68],"a":[16,32,82,95,108,151],"principled":[17,192],"mathematical":[18,71,193],"framework":[19,85,194],"leveraging":[21],"ordinal":[22,75,97,126,163],"preference":[23,62,76,210],"data.":[24],"When":[25],"annotators":[27],"provide":[28],"graded":[29],"preferences":[30,93,199],"on":[31,157],"Likert":[33,91,197],"scale":[34,92,198],"(e.g.,":[35],"significantly":[36],"better,":[37,38,40],"slightly":[39],"negligibly":[41],"better),":[42],"existing":[43,131,174],"methods":[44,133,176],"typically":[45],"apply":[46],"ad-hoc":[47,206],"heuristics,":[48],"such":[49],"as":[50,94],"margin":[51],"terms":[52],"or":[53,139,169],"scaling":[54,140],"factors,":[55],"to":[56,173,212],"loss":[57,103,111],"functions":[58,104],"derived":[59],"from":[60,105,148],"binary":[61,209],"like":[64],"Bradley-Terry.":[65],"These":[66],"an":[69,113],"underlying":[70],"model":[72,202],"how":[74],"data":[77,149],"generated.":[79],"We":[80,100],"present":[81],"theoretically":[83],"grounded":[84],"that":[86,122,134,161],"formulates":[87],"reward":[88,201],"discrete":[96],"regression":[98,164],"problem.":[99],"derive":[101],"two":[102],"this":[106],"formulation:":[107],"negative":[109],"log-likelihood":[110],"and":[112,184],"all-threshold":[114],"loss,":[115],"both":[116],"of":[117,128,208,217],"which":[118],"learn":[119],"threshold":[120],"parameters":[121,146],"naturally":[123],"capture":[124],"the":[125,190],"structure":[127],"preferences.":[129],"Unlike":[130],"heuristic":[132,175],"manually":[135],"specify":[136],"fixed":[137],"margins":[138],"weights,":[141],"our":[142,162],"approach":[143,165],"learns":[144],"these":[145],"directly":[147],"within":[150],"coherent":[152],"probabilistic":[153],"framework.":[154],"Experimental":[155],"results":[156],"multiple":[158],"benchmarks":[159],"demonstrate":[160],"consistently":[166],"achieves":[167],"competitive":[168],"superior":[170],"performance":[171],"compared":[172],"across":[177],"diverse":[178],"evaluation":[179],"categories":[180],"including":[181],"chat,":[182],"reasoning,":[183],"safety":[185],"tasks.":[186],"Our":[187],"work":[188],"provides":[189],"first":[191],"incorporating":[196],"into":[200],"training,":[203],"moving":[204],"beyond":[205],"modifications":[207],"enable":[213],"more":[214],"effective":[215],"utilization":[216],"fine-grained":[218],"feedback.":[220]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-03-05T00:00:00"}