{"id":"https://openalex.org/W7133521742","doi":"https://doi.org/10.48550/arxiv.2603.02546","title":"On Discriminative vs. Generative classifiers: Rethinking MLLMs for Action Understanding","display_name":"On Discriminative vs. Generative classifiers: Rethinking MLLMs for Action Understanding","publication_year":2026,"publication_date":"2026-03-03","ids":{"openalex":"https://openalex.org/W7133521742","doi":"https://doi.org/10.48550/arxiv.2603.02546"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.02546","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02546","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.02546","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5103856431","display_name":"Zhanzhong Pang","orcid":"https://orcid.org/0009-0008-8320-1727"},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Pang, Zhanzhong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055761154","display_name":"Dibyadip Chatterjee","orcid":"https://orcid.org/0000-0002-2651-3045"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chatterjee, Dibyadip","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128076645","display_name":"Fadime Sener","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sener, Fadime","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5121194827","display_name":"Angela Yao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yao, Angela","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5103856431"],"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":0.31520000100135803,"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":0.31520000100135803,"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/T10812","display_name":"Human Pose and Action Recognition","score":0.2766000032424927,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.06270000338554382,"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/discriminative-model","display_name":"Discriminative model","score":0.906000018119812},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.7515000104904175},{"id":"https://openalex.org/keywords/classifier","display_name":"Classifier (UML)","score":0.6656000018119812},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6193000078201294},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.5368000268936157},{"id":"https://openalex.org/keywords/ambiguity","display_name":"Ambiguity","score":0.5214999914169312}],"concepts":[{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.906000018119812},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.7515000104904175},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.7010999917984009},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.6656000018119812},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6248999834060669},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6193000078201294},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5703999996185303},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.5368000268936157},{"id":"https://openalex.org/C2780522230","wikidata":"https://www.wikidata.org/wiki/Q1140419","display_name":"Ambiguity","level":2,"score":0.5214999914169312},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.3912000060081482},{"id":"https://openalex.org/C112313634","wikidata":"https://www.wikidata.org/wiki/Q7886648","display_name":"Complement (music)","level":5,"score":0.3601999878883362},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3262999951839447},{"id":"https://openalex.org/C2987834672","wikidata":"https://www.wikidata.org/wiki/Q4677630","display_name":"Action recognition","level":3,"score":0.29030001163482666},{"id":"https://openalex.org/C106135958","wikidata":"https://www.wikidata.org/wiki/Q7291993","display_name":"Random subspace method","level":3,"score":0.2883000075817108}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.02546","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02546","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.02546","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.02546","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","score":0.7439162731170654,"display_name":"Reduced inequalities"}],"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],"advanced":[6],"open-world":[7],"action":[8,23,36,75,131,148],"understanding":[9,149],"and":[10,32,68,81,157,177],"can":[11,110],"be":[12],"adapted":[13],"as":[14,25],"generative":[15,67,96,108,160],"classifiers":[16,49,70,97],"for":[17,73,129],"closed-set":[18,74,130],"settings":[19],"by":[20],"autoregressively":[21],"generating":[22],"labels":[24,37],"text.":[26],"However,":[27],"this":[28,122],"approach":[29],"is":[30],"inefficient,":[31],"shared":[33],"subwords":[34],"across":[35,168],"introduce":[38],"semantic":[39],"overlap,":[40],"leading":[41,114],"to":[42,115],"ambiguity":[43],"in":[44],"generation.":[45],"In":[46],"contrast,":[47],"discriminative":[48,69,102,112],"learn":[50],"task-specific":[51],"representations":[52],"with":[53,71,101,141],"clear":[54],"decision":[55],"boundaries,":[56],"enabling":[57],"efficient":[58],"one-step":[59],"classification":[60],"without":[61],"autoregressive":[62],"decoding.":[63],"We":[64],"first":[65],"compare":[66],"MLLMs":[72],"understanding,":[76],"revealing":[77],"the":[78,84,88],"superior":[79],"accuracy":[80,156,175],"efficiency":[82,158],"of":[83],"latter.":[85],"To":[86,121],"bridge":[87],"performance":[89,99,117],"gap,":[90],"we":[91,105,124],"design":[92],"strategies":[93],"that":[94,107,152],"elevate":[95],"toward":[98],"comparable":[100],"ones.":[103],"Furthermore,":[104],"show":[106],"modeling":[109],"complement":[111],"classifiers,":[113],"better":[116],"while":[118],"preserving":[119,138],"efficiency.":[120],"end,":[123],"propose":[125],"Generation-Assisted":[126],"Discriminative~(GAD)":[127],"classifier":[128],"understanding.":[132],"GAD":[133,153],"operates":[134],"only":[135],"during":[136],"fine-tuning,":[137],"full":[139],"compatibility":[140],"MLLM":[142],"pretraining.":[143],"Extensive":[144],"experiments":[145],"on":[146,165,181],"temporal":[147],"benchmarks":[150],"demonstrate":[151],"improves":[154],"both":[155],"over":[159],"methods,":[161],"achieving":[162],"state-of-the-art":[163],"results":[164],"four":[166],"tasks":[167],"five":[169],"datasets,":[170],"including":[171],"an":[172],"average":[173],"2.5%":[174],"gain":[176],"3x":[178],"faster":[179],"inference":[180],"our":[182],"largest":[183],"COIN":[184],"benchmark.":[185]},"counts_by_year":[],"updated_date":"2026-03-05T07:36:02.291473","created_date":"2026-03-05T00:00:00"}