{"id":"https://openalex.org/W4389983704","doi":"https://doi.org/10.48550/arxiv.2312.10813","title":"Towards Efficient Vision-Language Tuning: More Information Density, More Generalizability","display_name":"Towards Efficient Vision-Language Tuning: More Information Density, More Generalizability","publication_year":2023,"publication_date":"2023-12-17","ids":{"openalex":"https://openalex.org/W4389983704","doi":"https://doi.org/10.48550/arxiv.2312.10813"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2312.10813","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2312.10813","pdf_url":"https://arxiv.org/pdf/2312.10813","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2312.10813","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5019176668","display_name":"Tianxiang Hao","orcid":"https://orcid.org/0000-0002-1952-6083","observed_orcids":["https://orcid.org/0000-0002-1952-6083"]},"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":"Hao, Tianxiang","raw_affiliation_strings":["School of Software , Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Software , Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5075344331","display_name":"Mengyao Lyu","orcid":"https://orcid.org/0000-0002-5404-4127","observed_orcids":["https://orcid.org/0000-0002-5404-4127"]},"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":"Lyu, Mengyao","raw_affiliation_strings":["School of Software , Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Software , Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101409147","display_name":"Hui Chen","orcid":"https://orcid.org/0000-0002-8010-8796","observed_orcids":["https://orcid.org/0000-0002-8010-8796"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Hui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051149140","display_name":"Sicheng Zhao","orcid":"https://orcid.org/0000-0001-5843-6411","observed_orcids":["https://orcid.org/0000-0001-5843-6411","https://orcid.org/0009-0001-1748-6663","https://orcid.org/0009-0002-5577-1913"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Sicheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Ding, Xiaohan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405256866","display_name":"ByteDance (China)","ror":"https://ror.org/04m8av638","country_code":"CN","type":"company","lineage":["https://openalex.org/I4405256866"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ding, Xiaohan","raw_affiliation_strings":["Bytedance"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Bytedance","institution_ids":["https://openalex.org/I4405256866"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5046605531","display_name":"Jungong Han","orcid":"https://orcid.org/0000-0003-4361-956X","observed_orcids":["https://orcid.org/0000-0003-4361-956X","https://orcid.org/0000-0002-2828-9905"]},"institutions":[{"id":"https://openalex.org/I91136226","display_name":"University of Sheffield","ror":"https://ror.org/05krs5044","country_code":"GB","type":"education","lineage":["https://openalex.org/I91136226"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Han, Jungong","raw_affiliation_strings":["Computer Science Department , University of Sheffied"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Computer Science Department , University of Sheffied","institution_ids":["https://openalex.org/I91136226"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5057732142","display_name":"Guiguang Ding","orcid":"https://orcid.org/0000-0003-0137-9975","observed_orcids":["https://orcid.org/0000-0003-0137-9975"]},"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":"Ding, Guiguang","raw_affiliation_strings":["School of Software , Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Software , Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":6,"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":0.8895999789237976,"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":0.8895999789237976,"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":0.1046999990940094,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.00139999995008111,"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/vision-language-models","display_name":"vision-language models","score":0.8220000267028809},{"id":"https://openalex.org/keywords/prompt-tuning","display_name":"prompt tuning","score":0.902999997138977},{"id":"https://openalex.org/keywords/information-density","display_name":"information density","score":0.7839999794960022},{"id":"https://openalex.org/keywords/parameter-efficient-fine-tuning","display_name":"parameter-efficient fine-tuning","score":0.968999981880188},{"id":"https://openalex.org/keywords/generalization","display_name":"generalization","score":0.7549999952316284},{"id":"https://openalex.org/keywords/data-scarce-settings","display_name":"data-scarce settings","score":0.7960000038146973},{"id":"https://openalex.org/keywords/stability-and-robustness","display_name":"robustness and stability","score":0.6060000061988831}],"concepts":[{"id":"https://openalex.org/C165464430","wikidata":"https://www.wikidata.org/wiki/Q1570441","display_name":"Parameterized complexity","level":2,"score":0.8113863468170166},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.797322154045105},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.7893632650375366},{"id":"https://openalex.org/C164226766","wikidata":"https://www.wikidata.org/wiki/Q7293202","display_name":"Rank (graph theory)","level":2,"score":0.6942834258079529},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.5972725749015808},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5162113308906555},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.49703648686408997},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.45541030168533325},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.4482099711894989},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.35303670167922974},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.32613834738731384},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.09743648767471313},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C120665830","wikidata":"https://www.wikidata.org/wiki/Q14620","display_name":"Optics","level":1,"score":0.0},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.0},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2312.10813","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2312.10813","pdf_url":"https://arxiv.org/pdf/2312.10813","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.2312.10813","doi":"https://doi.org/10.48550/arxiv.2312.10813","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2312.10813","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":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2312.10813","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2312.10813","pdf_url":"https://arxiv.org/pdf/2312.10813","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":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4389983704.pdf","grobid_xml":"https://content.openalex.org/works/W4389983704.grobid-xml"},"referenced_works_count":20,"referenced_works":["https://openalex.org/W24089286","https://openalex.org/W1846799578","https://openalex.org/W3002104146","https://openalex.org/W3152956381","https://openalex.org/W3168867926","https://openalex.org/W3174770825","https://openalex.org/W3185341429","https://openalex.org/W3205270560","https://openalex.org/W3215626407","https://openalex.org/W4225747866","https://openalex.org/W4281833441","https://openalex.org/W4281838252","https://openalex.org/W4310557340","https://openalex.org/W4320086030","https://openalex.org/W4368321032","https://openalex.org/W4386790226","https://openalex.org/W4402954393","https://openalex.org/W4403600613","https://openalex.org/W4405078778","https://openalex.org/W4414693319"],"related_works":["https://openalex.org/W4382603086","https://openalex.org/W4388266798","https://openalex.org/W4390874497","https://openalex.org/W7131112567","https://openalex.org/W4390873936","https://openalex.org/W4415547814","https://openalex.org/W4225747866","https://openalex.org/W4394593151","https://openalex.org/W6929478844","https://openalex.org/W4415707830"],"abstract_inverted_index":{"With":[0],"the":[1,10,43,50,99,107,111,115,133,139,153,160,184],"advancement":[2],"of":[3,45,52,110,135,155,177,188],"large":[4],"pre-trained":[5],"vision-language":[6],"models,":[7],"effectively":[8],"transferring":[9],"knowledge":[11],"embedded":[12],"within":[13],"these":[14],"foundational":[15],"models":[16],"to":[17,56,63,90,106,122,126,194],"downstream":[18,186],"tasks":[19,178],"has":[20],"become":[21],"a":[22,59,78,165,175],"pivotal":[23],"topic,":[24],"particularly":[25,145],"in":[26,147],"data-scarce":[27],"environments.":[28],"Recently,":[29],"parameter-efficient":[30],"fine-tuning":[31],"approaches,":[32],"especially":[33],"prompt":[34,46,112,190],"tuning,":[35,47],"have":[36],"garnered":[37],"considerable":[38],"attention.":[39],"To":[40],"better":[41],"understand":[42],"nature":[44],"we":[48],"propose":[49],"concept":[51],"``Information":[53],"Density''":[54],"(ID)":[55],"indicate":[57],"whether":[58],"matrix":[60],"strongly":[61],"belongs":[62],"certain":[64],"feature":[65,74,86],"spaces":[66,87],"rather":[67],"than":[68],"being":[69],"evenly":[70],"distributed":[71],"across":[72,84],"various":[73],"spaces.":[75],"We":[76],"suppose":[77],"higher":[79],"ID":[80],"with":[81,168],"strong":[82],"bias":[83],"some":[85],"naturally":[88],"leads":[89],"excellent":[91],"robustness":[92],"and":[93,138],"stability.":[94],"Our":[95],"research,":[96],"inspired":[97],"by":[98,164,192],"observation":[100],"that":[101],"generalizability":[102],"is":[103],"closely":[104],"linked":[105],"information":[108,124],"density":[109,125],"matrix,":[113],"introduces":[114],"Dense":[116],"Information":[117],"Prompt":[118],"(DIP).":[119],"DIP":[120,130,158,182],"aims":[121],"enhance":[123],"improve":[127],"generalization.":[128],"Furthermore,":[129],"significantly":[131],"reduces":[132],"number":[134],"tunable":[136],"parameters":[137],"requisite":[140],"storage":[141],"space,":[142],"making":[143],"it":[144],"advantageous":[146],"resource-constrained":[148],"settings.":[149],"Comprehensive":[150],"experiments":[151],"substantiate":[152],"superiority":[154],"DIP.":[156],"Notably,":[157],"surpasses":[159],"latest":[161],"state-of-the-art":[162],"methods":[163],"substantial":[166],"margin":[167],"an":[169],"exceptionally":[170],"small":[171],"parameter":[172],"count.":[173],"Across":[174],"range":[176],"spanning":[179],"11":[180],"datasets,":[181],"improves":[183],"average":[185],"accuracy":[187],"classic":[189],"tuning":[191],"up":[193],"5.76%":[195],"using":[196],"merely":[197],"0.5K":[198],"parameters.":[199]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":3}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2023-12-20T00:00:00"}