{"id":"https://openalex.org/W4392019983","doi":"https://doi.org/10.48550/arxiv.2402.12750","title":"Model Composition for Multimodal Large Language Models","display_name":"Model Composition for Multimodal Large Language Models","publication_year":2024,"publication_date":"2024-02-20","ids":{"openalex":"https://openalex.org/W4392019983","doi":"https://doi.org/10.48550/arxiv.2402.12750"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2402.12750","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.12750","pdf_url":"https://arxiv.org/pdf/2402.12750","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":null,"license_id":null,"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/2402.12750","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Chen, Chi","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Chi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100871849","display_name":"Yiyang Du","orcid":"https://orcid.org/0009-0004-8397-8246","observed_orcids":["https://orcid.org/0009-0004-8397-8246","https://orcid.org/0009-0007-0990-9981"]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Du, Yiyang","raw_affiliation_strings":["Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Fang, Zheng","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Fang, Zheng","raw_affiliation_strings":["Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102677951","display_name":"Ziyue Wang","orcid":"https://orcid.org/0009-0002-7041-0913","observed_orcids":["https://orcid.org/0009-0002-7041-0913","https://orcid.org/0009-0004-3176-5775","https://orcid.org/0009-0006-8239-8009"]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wang, Ziyue","raw_affiliation_strings":["Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114124897","display_name":"Fuwen Luo","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Luo, Fuwen","raw_affiliation_strings":["Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108144474","display_name":"Peng Li","orcid":"https://orcid.org/0000-0002-5876-2177","observed_orcids":["https://orcid.org/0000-0002-5876-2177"]},"institutions":[{"id":"https://openalex.org/I4391012619","display_name":"Shanghai Artificial Intelligence Laboratory","ror":"https://ror.org/03wkvpx79","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4391012619"]},{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Li, Peng","raw_affiliation_strings":["Institute for AI Industry Research (AIR) , Tsinghua University , Beijing , China","Shanghai Artificial Intelligence Laboratory , Shanghai , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute for AI Industry Research (AIR) , Tsinghua University , Beijing , China","institution_ids":["https://openalex.org/I99065089"]},{"raw_affiliation_string":"Shanghai Artificial Intelligence Laboratory , Shanghai , China","institution_ids":["https://openalex.org/I4391012619"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100452477","display_name":"Ming Yan","orcid":"https://orcid.org/0000-0003-4959-8878","observed_orcids":["https://orcid.org/0000-0003-4959-8878"]},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yan, Ming","raw_affiliation_strings":["Institute of Intelligent Computing , Alibaba Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Intelligent Computing , Alibaba Group","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100329266","display_name":"Ji Zhang","orcid":"https://orcid.org/0000-0002-3835-7975","observed_orcids":["https://orcid.org/0000-0002-3835-7975","https://orcid.org/0009-0008-7442-016X"]},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Ji","raw_affiliation_strings":["Institute of Intelligent Computing , Alibaba Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Intelligent Computing , Alibaba Group","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101488344","display_name":"Fei Huang","orcid":"https://orcid.org/0000-0002-3709-5053","observed_orcids":["https://orcid.org/0000-0002-3709-5053","https://orcid.org/0000-0002-2795-3114","https://orcid.org/0009-0005-7670-5255"]},"institutions":[{"id":"https://openalex.org/I45928872","display_name":"Alibaba Group (China)","ror":"https://ror.org/00k642b80","country_code":"CN","type":"company","lineage":["https://openalex.org/I45928872"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Huang, Fei","raw_affiliation_strings":["Institute of Intelligent Computing , Alibaba Group"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Intelligent Computing , Alibaba Group","institution_ids":["https://openalex.org/I45928872"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5046448314","display_name":"Maosong Sun","orcid":"https://orcid.org/0000-0002-6011-6115","observed_orcids":["https://orcid.org/0000-0002-6011-6115"]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Sun, Maosong","raw_affiliation_strings":["Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dept. of Comp. Sci. & Tech. , Institute for AI , Tsinghua University , Beijing , China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100760790","display_name":"Yang Liu","orcid":"https://orcid.org/0009-0005-3929-1583","observed_orcids":["https://orcid.org/0009-0005-3929-1583","https://orcid.org/0000-0002-0862-6046"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":3,"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":1.0,"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":1.0,"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":0.0,"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":0.0,"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/multimodal-large-language-models","display_name":"multimodal large language models","score":0.996999979019165},{"id":"https://openalex.org/keywords/model-composition","display_name":"model composition","score":0.9959999918937683},{"id":"https://openalex.org/keywords/multimodal-understanding","display_name":"multimodal understanding","score":0.7770000100135803},{"id":"https://openalex.org/keywords/benchmark","display_name":"benchmark","score":0.5529999732971191},{"id":"https://openalex.org/keywords/mcub","display_name":"MCUB","score":0.9929999709129333}],"concepts":[{"id":"https://openalex.org/C40231798","wikidata":"https://www.wikidata.org/wiki/Q1333743","display_name":"Composition (language)","level":2,"score":0.6385068893432617},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4758610129356384},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.42781782150268555},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.35070323944091797},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.12080013751983643}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2402.12750","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.12750","pdf_url":"https://arxiv.org/pdf/2402.12750","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2402.12750","doi":"https://doi.org/10.48550/arxiv.2402.12750","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2402.12750","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2402.12750","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.12750","pdf_url":"https://arxiv.org/pdf/2402.12750","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G1111291411","display_name":null,"funder_award_id":"62276152","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G7879634455","display_name":null,"funder_award_id":"61925601","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4392019983.pdf","grobid_xml":"https://content.openalex.org/works/W4392019983.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W2382290278","https://openalex.org/W2478288626","https://openalex.org/W4391913857","https://openalex.org/W2350741829","https://openalex.org/W2530322880"],"abstract_inverted_index":{"Recent":[0],"developments":[1],"in":[2,120],"Multimodal":[3],"Large":[4],"Language":[5],"Models":[6],"(MLLMs)":[7],"have":[8],"shown":[9],"rapid":[10],"progress,":[11],"moving":[12],"towards":[13],"the":[14,58,71,84,109,114],"goal":[15],"of":[16,61,75,86,131,164],"creating":[17],"versatile":[18,161],"MLLMs":[19,63,132],"that":[20,69,155],"understand":[21,134],"inputs":[22,135,166],"from":[23,136,167],"various":[24],"modalities.":[25,48,138,169],"However,":[26],"existing":[27,62],"methods":[28],"typically":[29],"rely":[30],"on":[31,140],"joint":[32],"training":[33],"with":[34],"paired":[35],"multimodal":[36,146],"instruction":[37],"data,":[38],"which":[39],"is":[40],"resource-intensive":[41],"and":[42,93,105,143],"challenging":[43],"to":[44,46,64,101,133],"extend":[45],"new":[47,55,67],"In":[49],"this":[50,87,121,141],"paper,":[51],"we":[52,98,123],"propose":[53,124],"a":[54,66,126,160],"paradigm":[56,88],"through":[57],"model":[59,68,115,156,162],"composition":[60,157],"create":[65,159],"retains":[70],"modal":[72],"understanding":[73,147],"capabilities":[74],"each":[76],"original":[77],"model.":[78],"Our":[79],"basic":[80],"implementation,":[81],"NaiveMC,":[82],"demonstrates":[83],"effectiveness":[85],"by":[89],"reusing":[90],"modality":[91],"encoders":[92],"merging":[94,110],"LLM":[95],"parameters.":[96],"Furthermore,":[97],"introduce":[99],"DAMC":[100],"address":[102],"parameter":[103],"interference":[104],"mismatch":[106],"issues":[107],"during":[108],"process,":[111],"thereby":[112],"enhancing":[113],"performance.":[116],"To":[117],"facilitate":[118],"research":[119],"area,":[122],"MCUB,":[125],"benchmark":[127,142],"for":[128],"assessing":[129],"ability":[130],"diverse":[137],"Experiments":[139],"four":[144],"other":[145],"tasks":[148],"show":[149],"significant":[150],"improvements":[151],"over":[152],"baselines,":[153],"proving":[154],"can":[158],"capable":[163],"processing":[165],"multiple":[168]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":1}],"updated_date":"2026-10-16T09:59:22.235173","created_date":"2025-10-10T00:00:00"}