{"id":"https://openalex.org/W4409368317","doi":"https://doi.org/10.1609/aaai.v39i4.32427","title":"What Kind of Visual Tokens Do We Need? Training-Free Visual Token Pruning for Multi-Modal Large Language Models from the Perspective of Graph","display_name":"What Kind of Visual Tokens Do We Need? Training-Free Visual Token Pruning for Multi-Modal Large Language Models from the Perspective of Graph","publication_year":2025,"publication_date":"2025-04-11","ids":{"openalex":"https://openalex.org/W4409368317","doi":"https://doi.org/10.1609/aaai.v39i4.32427"},"language":"en","primary_location":{"id":"doi:10.1609/aaai.v39i4.32427","doi":"https://doi.org/10.1609/aaai.v39i4.32427","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v39i4.32427","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/32427/34582","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":["jufo-2","norway-1"],"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"volume":null,"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://ojs.aaai.org/index.php/AAAI/article/download/32427/34582","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5053740363","display_name":"Yutao Jiang","orcid":"https://orcid.org/0009-0007-0276-6025","observed_orcids":["https://orcid.org/0009-0007-0276-6025"]},"institutions":[{"id":"https://openalex.org/I191208505","display_name":"Xiamen University","ror":"https://ror.org/00mcjh785","country_code":"CN","type":"education","lineage":["https://openalex.org/I191208505"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yutao Jiang","raw_affiliation_strings":["Xiamen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiamen University","institution_ids":["https://openalex.org/I191208505"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Qiong Wu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I191208505","display_name":"Xiamen University","ror":"https://ror.org/00mcjh785","country_code":"CN","type":"education","lineage":["https://openalex.org/I191208505"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qiong Wu","raw_affiliation_strings":["Xiamen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiamen University","institution_ids":["https://openalex.org/I191208505"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006955599","display_name":"Wenhao Lin","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I191208505","display_name":"Xiamen University","ror":"https://ror.org/00mcjh785","country_code":"CN","type":"education","lineage":["https://openalex.org/I191208505"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wenhao Lin","raw_affiliation_strings":["Xiamen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiamen University","institution_ids":["https://openalex.org/I191208505"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Wei Yu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I191208505","display_name":"Xiamen University","ror":"https://ror.org/00mcjh785","country_code":"CN","type":"education","lineage":["https://openalex.org/I191208505"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wei Yu","raw_affiliation_strings":["Xiamen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiamen University","institution_ids":["https://openalex.org/I191208505"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5091531008","display_name":"Yiyi Zhou","orcid":"https://orcid.org/0000-0002-5110-4526","observed_orcids":["https://orcid.org/0000-0002-5110-4526"]},"institutions":[{"id":"https://openalex.org/I191208505","display_name":"Xiamen University","ror":"https://ror.org/00mcjh785","country_code":"CN","type":"education","lineage":["https://openalex.org/I191208505"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yiyi Zhou","raw_affiliation_strings":["Xiamen University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Xiamen University","institution_ids":["https://openalex.org/I191208505"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I191208505"],"apc_list":null,"apc_paid":null,"fwci":4.7088,"has_fulltext":true,"cited_by_count":6,"citation_normalized_percentile":{"value":0.95273144,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":99},"biblio":{"volume":"39","issue":"4","first_page":"4075","last_page":"4083"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9970999956130981,"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":0.9970999956130981,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.0026000000070780516,"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/T11605","display_name":"Visual Attention and Saliency Detection","score":0.0,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/visual-token-pruning","display_name":"visual token pruning","score":0.8870000243186951},{"id":"https://openalex.org/keywords/multimodal-large-language-models","display_name":"multimodal large language models","score":0.9559999704360962},{"id":"https://openalex.org/keywords/training-free-pruning","display_name":"training-free pruning","score":0.9089999794960022},{"id":"https://openalex.org/keywords/visual-redundancy","display_name":"visual redundancy","score":0.6710000038146973},{"id":"https://openalex.org/keywords/semantic-similarity","display_name":"semantic similarity","score":0.5009999871253967},{"id":"https://openalex.org/keywords/llava-next","display_name":"LLaVA-NeXT","score":0.8399999737739563},{"id":"https://openalex.org/keywords/flops-reduction","display_name":"FLOPs reduction","score":0.6759999990463257},{"id":"https://openalex.org/keywords/visual-question-answering","display_name":"visual question answering","score":0.5630000233650208}],"concepts":[{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7329104542732239},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7111802101135254},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.5972670316696167},{"id":"https://openalex.org/C71139939","wikidata":"https://www.wikidata.org/wiki/Q910194","display_name":"Modal","level":2,"score":0.5635320544242859},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.4793206453323364},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.4760929346084595},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.4510301351547241},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.39393430948257446},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.35315781831741333},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.33454424142837524},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.07480338215827942},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","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/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C188027245","wikidata":"https://www.wikidata.org/wiki/Q750446","display_name":"Polymer chemistry","level":1,"score":0.0},{"id":"https://openalex.org/C6557445","wikidata":"https://www.wikidata.org/wiki/Q173113","display_name":"Agronomy","level":1,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1609/aaai.v39i4.32427","doi":"https://doi.org/10.1609/aaai.v39i4.32427","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v39i4.32427","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/32427/34582","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":["jufo-2","norway-1"],"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},{"id":"pmh:oai:arXiv.org:2501.02268","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2501.02268","pdf_url":"https://arxiv.org/pdf/2501.02268","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2501.02268","doi":"https://doi.org/10.48550/arxiv.2501.02268","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2501.02268","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.1609/aaai.v39i4.32427","doi":"https://doi.org/10.1609/aaai.v39i4.32427","is_oa":true,"landing_page_url":"https://doi.org/10.1609/aaai.v39i4.32427","pdf_url":"https://ojs.aaai.org/index.php/AAAI/article/download/32427/34582","source":{"id":"https://openalex.org/S4210191458","display_name":"Proceedings of the AAAI Conference on Artificial Intelligence","issn_l":"2159-5399","issn":["2159-5399","2374-3468"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":["jufo-2","norway-1"],"host_organization":"https://openalex.org/P4310320058","host_organization_name":"Association for the Advancement of Artificial Intelligence","host_organization_lineage":["https://openalex.org/P4310320058"],"host_organization_lineage_names":["Association for the Advancement of Artificial Intelligence"],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the AAAI Conference on Artificial Intelligence","raw_type":"journal-article"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G1271955826","display_name":"\u57fa\u4e8e\u5f31\u76d1\u7763\u5b66\u4e60\u7684\u673a\u8f7d\u5730/\u6d77\u5c0f\u76ee\u6807\u68c0\u6d4b\u7814\u7a76","funder_award_id":"62072387","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G1300220747","display_name":null,"funder_award_id":"62025603","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G1787173488","display_name":"\u9762\u5411\u6df1\u5ea6\u5b66\u4e60\u7684\u5bf9\u6297\u653b\u9632\u6280\u672f\u7814\u7a76","funder_award_id":"62072389","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G187639494","display_name":null,"funder_award_id":"No. 62025603","funder_id":"https://openalex.org/F4320336125","funder_display_name":"National Science Fund for Distinguished Young Scholars"},{"id":"https://openalex.org/G2564228358","display_name":null,"funder_award_id":"No. 62025603","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G2785666877","display_name":null,"funder_award_id":"62002305","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G3181244293","display_name":null,"funder_award_id":"2022J06001","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G3568142893","display_name":null,"funder_award_id":"62272401","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G363226264","display_name":null,"funder_award_id":"U23A20383","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G3794798619","display_name":null,"funder_award_id":"U21A20472","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G5284885812","display_name":null,"funder_award_id":"62176222","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G6075090814","display_name":null,"funder_award_id":"62025603","funder_id":"https://openalex.org/F4320336125","funder_display_name":"National Science Fund for Distinguished Young Scholars"},{"id":"https://openalex.org/G6181615979","display_name":null,"funder_award_id":"2021J06003","funder_id":"https://openalex.org/F4320321878","funder_display_name":"Natural Science Foundation of Fujian Province"},{"id":"https://openalex.org/G6727647225","display_name":null,"funder_award_id":"62176226","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G717600273","display_name":null,"funder_award_id":"U22B2051","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G7220565969","display_name":null,"funder_award_id":"2022J06001","funder_id":"https://openalex.org/F4320321878","funder_display_name":"Natural Science Foundation of Fujian Province"},{"id":"https://openalex.org/G7272499820","display_name":null,"funder_award_id":"U21B2037","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G7779606772","display_name":null,"funder_award_id":"62176223","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G8406165135","display_name":"\u9762\u5411\u56fe\u654f\u611f\u6df1\u5ea6\u5b66\u4e60\u7684\u7ed3\u6784\u5316\u8d85\u56fe\u7814\u7a76","funder_award_id":"62072386","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"},{"id":"https://openalex.org/F4320321878","display_name":"Natural Science Foundation of Fujian Province","ror":null},{"id":"https://openalex.org/F4320336125","display_name":"National Science Fund for Distinguished Young Scholars","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4409368317.pdf","grobid_xml":"https://content.openalex.org/works/W4409368317.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4388335561","https://openalex.org/W2970530566","https://openalex.org/W4288261899","https://openalex.org/W4307309205","https://openalex.org/W2967478618","https://openalex.org/W4385009901","https://openalex.org/W4385572700","https://openalex.org/W2997152889","https://openalex.org/W230091440","https://openalex.org/W4285141722"],"abstract_inverted_index":{"Recent":[0],"Multimodal":[1],"Large":[2],"Language":[3],"Models(MLLMs)":[4],"often":[5],"use":[6],"a":[7,64,124,134],"large":[8],"number":[9],"of":[10,33,56,136,166],"visual":[11,16,24,34,69,78],"tokens":[12,35,47,79,104],"to":[13,19,123],"compensate":[14],"their":[15,84,88],"shortcoming,":[17],"leading":[18],"excessive":[20],"computation":[21,147],"and":[22,40,45,82,100,129,156,170,175],"obvious":[23],"redundancy.":[25],"In":[26,74],"this":[27,60],"paper,":[28],"we":[29,62,120],"investigate":[30],"what":[31],"kind":[32],"are":[36,48,107],"needed":[37],"for":[38,50,109],"MLLMs,":[39,110],"reveal":[41],"that":[42,142],"both":[43,154],"foreground":[44],"background":[46],"critical":[49],"MLLMs":[51],"given":[52],"the":[53,92,101],"varying":[54],"difficulties":[55],"examples.":[57],"Based":[58],"on":[59,87,133,153,168],"observation,":[61],"propose":[63],"graph-based":[65],"method":[66],"towards":[67],"training-free":[68],"token":[70],"pruning,":[71],"termed":[72],"G-Prune.":[73],"particular,":[75],"G-Prune":[76,143,161],"regards":[77],"as":[80],"nodes,":[81],"construct":[83],"connections":[85],"based":[86],"semantic":[89],"similarities.":[90],"Afterwards,":[91],"information":[93],"flow":[94],"is":[95],"propagated":[96],"via":[97],"weighted":[98],"links,":[99],"most":[102],"important":[103],"after":[105],"iterations":[106],"kept":[108],"which":[111],"can":[112,144,162],"be":[113],"front":[114],"or":[115],"background.":[116],"To":[117],"validate":[118],"G-Prune,":[119],"apply":[121],"it":[122],"recent":[125],"MLLM":[126],"called":[127],"LLaVA-NeXT,":[128],"conduct":[130],"extensive":[131],"experiments":[132],"set":[135],"benchmarks.":[137],"The":[138],"experiment":[139],"results":[140],"show":[141],"greatly":[145],"reduce":[146,163],"overhead":[148],"while":[149],"retaining":[150],"high":[151],"performance":[152],"coarse-":[155],"fine-grained":[157],"tasks.":[158],"For":[159],"instance,":[160],"63.57%":[164],"FLOPs":[165],"LLaVA-NeXT":[167],"VQA2.0":[169],"TextVQA":[171],"with":[172],"only":[173],"0.95%":[174],"2.34%":[176],"accuracy":[177],"drops,":[178],"respectively.":[179]},"counts_by_year":[{"year":2026,"cited_by_count":6}],"updated_date":"2026-10-14T16:39:01.820549","created_date":"2025-10-10T00:00:00"}