{"id":"https://openalex.org/W7166865689","doi":"https://doi.org/10.18653/v1/2026.acl-long.601","title":"Vista-LLM: Decoupled Query-Guided Visual Token Pruning for Efficient Long-Video Large Language Models","display_name":"Vista-LLM: Decoupled Query-Guided Visual Token Pruning for Efficient Long-Video Large Language Models","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166865689","doi":"https://doi.org/10.18653/v1/2026.acl-long.601"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.acl-long.601","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.601","pdf_url":"https://aclanthology.org/2026.acl-long.601.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.acl-long.601.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139841212","display_name":"Zhenyu Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhenyu Li","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139818734","display_name":"Zuchao Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zuchao Li","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139772505","display_name":"Ping Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ping Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139747431","display_name":"Lefei Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lefei Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5074448137","display_name":"Haojun Ai","orcid":"https://orcid.org/0000-0002-4172-5070"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Haojun Ai","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.87808187,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"13171","last_page":"13187"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.8593000173568726,"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":0.8593000173568726,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.08259999752044678,"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":0.009700000286102295,"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/pruning","display_name":"Pruning","score":0.4523000121116638},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4074000120162964},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.3176000118255615},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.3057999908924103},{"id":"https://openalex.org/keywords/visualization","display_name":"Visualization","score":0.3050000071525574},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.27300000190734863}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6748999953269958},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5543000102043152},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.4523000121116638},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4074000120162964},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.32429999113082886},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.3176000118255615},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3059999942779541},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.3057999908924103},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3050000071525574},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.3003000020980835},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.27639999985694885},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.27300000190734863},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2565000057220459},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.25380000472068787},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25189998745918274},{"id":"https://openalex.org/C2780878386","wikidata":"https://www.wikidata.org/wiki/Q1659648","display_name":"Visual language","level":2,"score":0.25}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.601","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.601","pdf_url":"https://aclanthology.org/2026.acl-long.601.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.601","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.601","pdf_url":"https://aclanthology.org/2026.acl-long.601.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166865689.pdf","grobid_xml":"https://content.openalex.org/works/W7166865689.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Long-video":[0],"understanding":[1],"is":[2],"bottlenecked":[3],"by":[4,106],"the":[5],"high":[6],"cost":[7],"of":[8,115],"processing":[9],"massive":[10],"visual":[11,39,104,122],"tokens.Current":[12],"reduction":[13],"strategies":[14],"often":[15],"rely":[16],"on":[17,88,100,118],"static":[18],"allocation":[19],"or":[20],"inefficient":[21],"in-network":[22],"selection":[23],"that":[24],"disrupts":[25],"optimized":[26],"attention":[27],"kernels.In":[28],"this":[29],"paper,":[30],"we":[31],"introduce":[32],"Vista-LLM,":[33],"a":[34,59,71,95],"decoupled":[35],"framework":[36],"for":[37,66,75],"query-guided":[38],"token":[40],"pruning.By":[41],"filtering":[42,121],"redundancy":[43],"prior":[44],"to":[45,83],"inference":[46,110],"with":[47,54],"minimal":[48],"overhead,":[49],"Vista-LLM":[50,102],"ensures":[51],"full":[52],"compatibility":[53],"Flash":[55],"Attention.Our":[56],"method":[57],"employs":[58],"coarse-tofine":[60],"pipeline:":[61],"(1)":[62],"Query-Guided":[63],"Dynamic":[64],"Budgeting":[65],"adaptive":[67],"temporal":[68],"allocation;":[69],"(2)":[70],"lightweight":[72],"Semantic":[73],"Scout":[74],"fine-grained,":[76],"query-specific":[77],"selection;":[78],"and":[79,92,108],"(3)":[80],"Structure-Aware":[81],"Compensation":[82],"preserve":[84],"global":[85],"context.Extensive":[86],"experiments":[87],"benchmarks":[89],"like":[90],"Video-MME":[91],"MLVU":[93],"demonstrate":[94],"significantly":[96],"improved":[97],"Pareto":[98],"frontier.Notably,":[99],"LLaVA-OneVision,":[101],"reduces":[103],"tokens":[105],"90%":[107],"accelerates":[109],"while":[111],"retaining":[112],"over":[113],"98%":[114],"baseline":[116],"performance":[117],"average,":[119],"effectively":[120],"noise.":[123]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}