{"id":"https://openalex.org/W4414810689","doi":"https://doi.org/10.48550/arxiv.2510.01146","title":"mR3: Multilingual Rubric-Agnostic Reward Reasoning Models","display_name":"mR3: Multilingual Rubric-Agnostic Reward Reasoning Models","publication_year":2025,"publication_date":"2025-10-01","ids":{"openalex":"https://openalex.org/W4414810689","doi":"https://doi.org/10.48550/arxiv.2510.01146"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2510.01146","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.01146","pdf_url":"https://arxiv.org/pdf/2510.01146","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/2510.01146","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5093876039","display_name":"David Anugraha","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Anugraha, David","raw_affiliation_strings":["Stanford University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Hung, Shou-Yi","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Hung, Shou-Yi","raw_affiliation_strings":["University of Toronto"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Toronto","institution_ids":["https://openalex.org/I185261750"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076315983","display_name":"Zilu Tang","orcid":"https://orcid.org/0000-0002-6504-8908","observed_orcids":["https://orcid.org/0000-0002-6504-8908"]},"institutions":[{"id":"https://openalex.org/I111088046","display_name":"Boston University","ror":"https://ror.org/05qwgg493","country_code":"US","type":"education","lineage":["https://openalex.org/I111088046"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Tang, Zilu","raw_affiliation_strings":["Boston University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Boston University","institution_ids":["https://openalex.org/I111088046"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078275178","display_name":"Annie En-Shiun Lee","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lee, Annie En-Shiun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027174211","display_name":"Derry Tanti Wijaya","orcid":"https://orcid.org/0000-0002-0848-4703","observed_orcids":["https://orcid.org/0000-0002-0848-4703"]},"institutions":[{"id":"https://openalex.org/I111088046","display_name":"Boston University","ror":"https://ror.org/05qwgg493","country_code":"US","type":"education","lineage":["https://openalex.org/I111088046"]},{"id":"https://openalex.org/I4405262958","display_name":"Monash University, Indonesia","ror":"https://ror.org/00m57t258","country_code":"ID","type":"education","lineage":["https://openalex.org/I4405262958","https://openalex.org/I56590836"]}],"countries":["ID","US"],"is_corresponding":false,"raw_author_name":"Wijaya, Derry Tanti","raw_affiliation_strings":["Boston University","Monash University Indonesia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Boston University","institution_ids":["https://openalex.org/I111088046"]},{"raw_affiliation_string":"Monash University Indonesia","institution_ids":["https://openalex.org/I4405262958"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5085516032","display_name":"Genta Indra Winata","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I1305444813","display_name":"Capital One (United States)","ror":"https://ror.org/00svp7168","country_code":"US","type":"company","lineage":["https://openalex.org/I1305444813"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Winata, Genta Indra","raw_affiliation_strings":["Capital One"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Capital One","institution_ids":["https://openalex.org/I1305444813"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.6542999744415283,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.6542999744415283,"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/T10028","display_name":"Topic Modeling","score":0.2549000084400177,"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/T10286","display_name":"Information Retrieval and Search Behavior","score":0.006500000134110451,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/reward-modeling","display_name":"reward modeling","score":0.843999981880188},{"id":"https://openalex.org/keywords/automated-assessment","display_name":"automated assessment","score":0.6110000014305115},{"id":"https://openalex.org/keywords/training-sample-selection","display_name":"training data selection","score":0.6430000066757202},{"id":"https://openalex.org/keywords/curriculum-learning","display_name":"curriculum learning","score":0.7350000143051147},{"id":"https://openalex.org/keywords/low-resource-languages","display_name":"low-resource languages","score":0.746999979019165},{"id":"https://openalex.org/keywords/reasoning-traces","display_name":"reasoning traces","score":0.8999999761581421},{"id":"https://openalex.org/keywords/human-evaluation","display_name":"human evaluation","score":0.32100000977516174}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7864999771118164},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6187000274658203},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5439000129699707},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.49559998512268066},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.46560001373291016},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4551999866962433},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4332999885082245},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41110000014305115},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.38530001044273376},{"id":"https://openalex.org/C2781249084","wikidata":"https://www.wikidata.org/wiki/Q908656","display_name":"Preference","level":2,"score":0.28999999165534973},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.28349998593330383},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.2791999876499176},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.26899999380111694},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.2624000012874603},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2605000138282776},{"id":"https://openalex.org/C24756922","wikidata":"https://www.wikidata.org/wiki/Q1757694","display_name":"Data quality","level":3,"score":0.25429999828338623},{"id":"https://openalex.org/C2983448237","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Language understanding","level":2,"score":0.2540999948978424}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2510.01146","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.01146","pdf_url":"https://arxiv.org/pdf/2510.01146","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.2510.01146","doi":"https://doi.org/10.48550/arxiv.2510.01146","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2510.01146","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:2510.01146","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2510.01146","pdf_url":"https://arxiv.org/pdf/2510.01146","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":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4414810689.pdf","grobid_xml":"https://content.openalex.org/works/W4414810689.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4404336932","https://openalex.org/W7212027601","https://openalex.org/W4417287612","https://openalex.org/W7155452057","https://openalex.org/W4416118041","https://openalex.org/W7166873924","https://openalex.org/W7125878423","https://openalex.org/W7134860561","https://openalex.org/W7140132325","https://openalex.org/W7167225266"],"abstract_inverted_index":{"Evaluation":[0],"using":[1],"Large":[2],"Language":[3],"Model":[4],"(LLM)":[5],"judges":[6],"has":[7],"been":[8],"widely":[9],"adopted":[10],"in":[11,65,98,141],"English":[12],"and":[13,31,77,86,124,145,153,185],"shown":[14],"to":[15,28,68,82,121],"be":[16],"effective":[17,37,84],"for":[18,40,80,89,96,172],"automatic":[19],"evaluation.":[20],"However,":[21],"their":[22],"performance":[23,106],"does":[24],"not":[25],"generalize":[26],"well":[27],"non-English":[29],"settings,":[30],"it":[32],"remains":[33],"unclear":[34],"what":[35],"constitutes":[36],"multilingual":[38,108],"training":[39,81],"such":[41],"judges.":[42],"In":[43],"this":[44],"paper,":[45],"we":[46,135],"introduce":[47],"mR3,":[48],"a":[49,72],"massively":[50],"multilingual,":[51],"rubric-agnostic":[52],"reward":[53,66,92,109],"reasoning":[54,97,151,168],"model":[55,110],"trained":[56],"on":[57,107],"72":[58],"languages,":[59,164],"achieving":[60],"the":[61,99,137,147],"broadest":[62],"language":[63],"coverage":[64],"modeling":[67],"date.":[69],"We":[70],"present":[71],"comprehensive":[73],"study":[74],"of":[75,139,149],"data":[76,87],"curriculum":[78],"selection":[79],"identify":[83],"strategies":[85],"sources":[88],"building":[90],"high-quality":[91],"models,":[93,183],"including":[94,171],"support":[95],"target":[100],"language.":[101],"Our":[102,182],"approach":[103],"attains":[104],"state-of-the-art":[105],"benchmarks,":[111],"surpassing":[112],"much":[113],"larger":[114],"models":[115],"(i.e.,":[116],"GPT-OSS-120B)":[117],"while":[118],"being":[119],"up":[120],"9x":[122],"smaller,":[123],"its":[125,150],"effectiveness":[126,138],"is":[127,169],"further":[128],"confirmed":[129],"through":[130,156],"extensive":[131],"ablation":[132],"studies.":[133],"Finally,":[134],"demonstrate":[136],"mR3":[140,166],"off-policy":[142],"preference":[143],"optimization":[144],"validate":[146],"quality":[148],"traces":[152],"rubric-based":[154],"evaluations":[155],"human":[157],"studies":[158],"with":[159],"20":[160],"annotators":[161],"across":[162],"12":[163],"where":[165],"models'":[167],"preferred,":[170],"extremely":[173],"low-resource":[174],"languages":[175],"that":[176],"are":[177,187],"entirely":[178],"unseen":[179],"during":[180],"training.":[181],"data,":[184],"code":[186],"available":[188],"as":[189],"open":[190],"source":[191],"at":[192],"https://github.com/rubricreward/mr3.":[193]},"counts_by_year":[],"updated_date":"2026-09-30T09:27:27.532063","created_date":"2025-10-10T00:00:00"}