{"id":"https://openalex.org/W7135173903","doi":"https://doi.org/10.48550/arxiv.2603.11665","title":"Multi-Task Reinforcement Learning for Enhanced Multimodal LLM-as-a-Judge","display_name":"Multi-Task Reinforcement Learning for Enhanced Multimodal LLM-as-a-Judge","publication_year":2026,"publication_date":"2026-03-12","ids":{"openalex":"https://openalex.org/W7135173903","doi":"https://doi.org/10.48550/arxiv.2603.11665"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.11665","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.11665","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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.2603.11665","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5129049857","display_name":"Junjie Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Junjie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5020981458","display_name":"Xuan Kan","orcid":"https://orcid.org/0000-0003-0039-8314"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kan, Xuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129000911","display_name":"Zihao He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Zihao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129009094","display_name":"Shunwen Tan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tan, Shunwen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5129061527","display_name":"Bo Pan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pan, Bo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5128986669","display_name":"Kaitai Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Kaitai","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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.8080000281333923,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.8080000281333923,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.08500000089406967,"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":0.015200000256299973,"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/generalization","display_name":"Generalization","score":0.7925000190734863},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.7002999782562256},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.6692000031471252},{"id":"https://openalex.org/keywords/reinforcement","display_name":"Reinforcement","score":0.26910001039505005},{"id":"https://openalex.org/keywords/task-analysis","display_name":"Task analysis","score":0.24770000576972961}],"concepts":[{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.7925000190734863},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.737500011920929},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7002999782562256},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.6692000031471252},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6590999960899353},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5539000034332275},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.26910001039505005},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.24770000576972961},{"id":"https://openalex.org/C117765406","wikidata":"https://www.wikidata.org/wiki/Q5362437","display_name":"Generalization error","level":3,"score":0.24050000309944153},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.23669999837875366}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.11665","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.11665","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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.2603.11665","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.11665","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":[{"id":"https://metadata.un.org/sdg/16","score":0.522522509098053,"display_name":"Peace, Justice and strong institutions"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Multimodal":[0],"Large":[1],"Language":[2],"Models":[3],"(MLLMs)":[4],"have":[5],"been":[6],"widely":[7],"adopted":[8],"as":[9],"MLLM-as-a-Judges":[10],"due":[11],"to":[12,35,37],"their":[13],"strong":[14,81,87],"alignment":[15],"with":[16,95],"human":[17,96],"judgment":[18,91],"across":[19,68],"various":[20],"visual":[21],"tasks.":[22],"However,":[23],"most":[24],"existing":[25],"judge":[26,66],"models":[27],"are":[28],"optimized":[29],"for":[30,45,57],"single-task":[31],"scenarios":[32],"and":[33,93],"struggle":[34],"generalize":[36],"diverse":[38],"contexts,":[39],"which":[40],"is":[41],"a":[42,60],"critical":[43],"requirement":[44],"reliable":[46],"evaluation.":[47],"To":[48],"address":[49],"this":[50],"limitation,":[51],"we":[52],"propose":[53],"Multi-Task":[54],"Reinforcement":[55],"Learning":[56],"MLLM-as-a-Judge":[58],"(MT-RL-Judge),":[59],"framework":[61],"that":[62,84],"jointly":[63],"optimizes":[64],"the":[65,72],"model":[67],"multiple":[69],"tasks,":[70,106],"leveraging":[71],"generalization":[73,103],"capabilities":[74],"of":[75],"RL.":[76],"Experimental":[77],"results":[78],"against":[79],"several":[80],"baselines":[82,88],"demonstrate":[83],"MT-RL-Judge":[85],"outperforms":[86],"in":[89],"both":[90],"consistency":[92],"correlation":[94],"preferences.":[97],"Furthermore,":[98],"our":[99],"approach":[100],"exhibits":[101],"robust":[102],"on":[104],"out-of-distribution":[105],"further":[107],"validating":[108],"its":[109],"effectiveness.":[110]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-03-14T00:00:00"}