{"id":"https://openalex.org/W7161979934","doi":"https://doi.org/10.48550/arxiv.2605.21059","title":"Multimodal LLMs under Pairwise Modalities","display_name":"Multimodal LLMs under Pairwise Modalities","publication_year":2026,"publication_date":"2026-05-20","ids":{"openalex":"https://openalex.org/W7161979934","doi":"https://doi.org/10.48550/arxiv.2605.21059"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.21059","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21059","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.2605.21059","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5136632444","display_name":"Li, Yan, 1978-","orcid":"https://orcid.org/0000-0003-1146-5578"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063510660","display_name":"Deng Yun-long","orcid":"https://orcid.org/0000-0003-2544-8146"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Deng, Yunlong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103266231","display_name":"Yuewen Sun","orcid":"https://orcid.org/0000-0001-7534-1297"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Yuewen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5012997528","display_name":"Gongxu Luo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Gongxu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136630040","display_name":"Kun Zhang","orcid":"https://orcid.org/0000-0003-4617-6258"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Kun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136695012","display_name":"Guangyi Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Guangyi","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":0.6328999996185303,"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":0.6328999996185303,"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/T10028","display_name":"Topic Modeling","score":0.10440000146627426,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.08229999989271164,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/modalities","display_name":"Modalities","score":0.8596000075340271},{"id":"https://openalex.org/keywords/pairwise-comparison","display_name":"Pairwise comparison","score":0.6784999966621399},{"id":"https://openalex.org/keywords/multimodal-learning","display_name":"Multimodal learning","score":0.6449999809265137},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.6150000095367432},{"id":"https://openalex.org/keywords/autoencoder","display_name":"Autoencoder","score":0.5135999917984009},{"id":"https://openalex.org/keywords/modality","display_name":"Modality (human\u2013computer interaction)","score":0.483599990606308},{"id":"https://openalex.org/keywords/feature-learning","display_name":"Feature learning","score":0.478300005197525},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.4410000145435333}],"concepts":[{"id":"https://openalex.org/C2779903281","wikidata":"https://www.wikidata.org/wiki/Q6888026","display_name":"Modalities","level":2,"score":0.8596000075340271},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7074000239372253},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6845999956130981},{"id":"https://openalex.org/C184898388","wikidata":"https://www.wikidata.org/wiki/Q1435712","display_name":"Pairwise comparison","level":2,"score":0.6784999966621399},{"id":"https://openalex.org/C2780660688","wikidata":"https://www.wikidata.org/wiki/Q25052564","display_name":"Multimodal learning","level":2,"score":0.6449999809265137},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.6150000095367432},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5408999919891357},{"id":"https://openalex.org/C101738243","wikidata":"https://www.wikidata.org/wiki/Q786435","display_name":"Autoencoder","level":3,"score":0.5135999917984009},{"id":"https://openalex.org/C2780226545","wikidata":"https://www.wikidata.org/wiki/Q6888030","display_name":"Modality (human\u2013computer interaction)","level":2,"score":0.483599990606308},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.478300005197525},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.4410000145435333},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.4341000020503998},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4041999876499176},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3517000079154968},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.35089999437332153},{"id":"https://openalex.org/C64543145","wikidata":"https://www.wikidata.org/wiki/Q162942","display_name":"Intersection (aeronautics)","level":2,"score":0.3142000138759613},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.29840001463890076},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2896000146865845},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.2802000045776367},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.27639999985694885},{"id":"https://openalex.org/C2778012447","wikidata":"https://www.wikidata.org/wiki/Q1034415","display_name":"Scope (computer science)","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C51167844","wikidata":"https://www.wikidata.org/wiki/Q4422623","display_name":"Latent variable","level":2,"score":0.2526000142097473}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.21059","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21059","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.2605.21059","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.21059","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":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Despite":[0],"the":[1,52,65,69,112,117,138,154,161,164],"impressive":[2],"results":[3],"achieved":[4],"by":[5,42,123,142,177,196],"multimodal":[6,18,55],"large":[7],"language":[8],"models":[9],"(MLLMs),":[10],"their":[11],"training":[12,40],"typically":[13],"relies":[14],"on":[15,79],"jointly":[16],"curated":[17],"data,":[19],"requiring":[20],"substantial":[21],"human":[22],"effort":[23],"to":[24,167],"construct":[25],"multi-way":[26],"aligned":[27,199],"datasets":[28],"and":[29,107,127,145,171,183,193],"thereby":[30],"limiting":[31],"scalability":[32],"across":[33,92,121],"domains.":[34],"In":[35,149],"this":[36,80],"work,":[37],"we":[38,58,82,115,152],"explore":[39],"MLLMs":[41,188],"only":[43,74,95],"leveraging":[44],"multiple":[45],"paired":[46],"modalities":[47,93,122,159,166,185],"as":[48],"a":[49,61,84],"surrogate":[50],"for":[51,88],"full":[53],"joint":[54],"distribution.":[56],"Specifically,":[57,110],"first":[59,113],"provide":[60],"theoretical":[62],"analysis":[63],"of":[64,101,156,163],"conditions":[66],"under":[67],"which":[68],"representations":[70,91],"are":[71],"identifiable":[72],"with":[73,160,189],"observing":[75],"pairwise":[76,96],"modalities.":[77],"Building":[78],"analysis,":[81],"propose":[83],"representation":[85,105,201],"learning":[86,140,197],"framework":[87,99],"aligning":[89,144],"latent":[90,104,119,147,200],"using":[94],"data.":[97],"The":[98],"consists":[100],"two":[102],"stages:":[103],"alignment":[106],"cross-modal":[108,169,207],"recomposition.":[109],"in":[111,137],"stage,":[114],"learn":[116],"shared":[118],"space":[120],"both":[124],"self-modal":[125],"reconstruction":[126],"pair-wise":[128],"contrastive":[129,139],"learning.":[130],"We":[131,173],"also":[132],"incorporate":[133],"an":[134,198],"inductive":[135],"bias":[136],"process":[141],"partially":[143],"minimal":[146],"specification.":[148],"stage":[150],"two,":[151],"integrate":[153],"encoder":[155],"newly":[157,178],"introduced":[158],"decoders":[162],"pre-trained":[165,187],"facilitate":[168],"transfer":[170],"generation.":[172],"evaluate":[174],"our":[175,203],"method":[176],"adding":[179],"3D":[180],"point":[181],"clouds":[182],"tactile":[184],"into":[186],"three":[190],"modality":[191],"pairs":[192],"show":[194],"that,":[195],"space,":[202],"model":[204],"achieves":[205],"strong":[206],"performance.":[208]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-22T00:00:00"}