{"id":"https://openalex.org/W7143304641","doi":"https://doi.org/10.48550/arxiv.2603.26041","title":"Rethinking Token Pruning for Historical Screenshots in GUI Visual Agents: Semantic, Spatial, and Temporal Perspectives","display_name":"Rethinking Token Pruning for Historical Screenshots in GUI Visual Agents: Semantic, Spatial, and Temporal Perspectives","publication_year":2026,"publication_date":"2026-03-27","ids":{"openalex":"https://openalex.org/W7143304641","doi":"https://doi.org/10.48550/arxiv.2603.26041"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.26041","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.26041","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.26041","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5130928274","display_name":"Daiqiang Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Daiqiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5053945360","display_name":"Zihao Pan","orcid":"https://orcid.org/0000-0001-8882-4589"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pan, Zihao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130910329","display_name":"Zeyu Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Zeyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130956479","display_name":"Ronghao Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Ronghao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130970385","display_name":"Huacan Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Huacan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130914100","display_name":"Honggang Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Honggang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5130934055","display_name":"Haiyun Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiang, Haiyun","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":0.7508000135421753,"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":0.7508000135421753,"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/T10709","display_name":"Social Robot Interaction and HRI","score":0.042399998754262924,"subfield":{"id":"https://openalex.org/subfields/3207","display_name":"Social Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10789","display_name":"Interactive and Immersive Displays","score":0.020099999383091927,"subfield":{"id":"https://openalex.org/subfields/1709","display_name":"Human-Computer Interaction"},"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":0.8009999990463257},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.7297000288963318},{"id":"https://openalex.org/keywords/partition","display_name":"Partition (number theory)","score":0.4020000100135803},{"id":"https://openalex.org/keywords/graphical-user-interface","display_name":"Graphical user interface","score":0.3682999908924103},{"id":"https://openalex.org/keywords/graphical-model","display_name":"Graphical model","score":0.32420000433921814},{"id":"https://openalex.org/keywords/visualization","display_name":"Visualization","score":0.3073999881744385}],"concepts":[{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.8009999990463257},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7901999950408936},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.7297000288963318},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4966999888420105},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.4415000081062317},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.43720000982284546},{"id":"https://openalex.org/C42812","wikidata":"https://www.wikidata.org/wiki/Q1082910","display_name":"Partition (number theory)","level":2,"score":0.4020000100135803},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3781999945640564},{"id":"https://openalex.org/C37789001","wikidata":"https://www.wikidata.org/wiki/Q782543","display_name":"Graphical user interface","level":2,"score":0.3682999908924103},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.32839998602867126},{"id":"https://openalex.org/C155846161","wikidata":"https://www.wikidata.org/wiki/Q1143367","display_name":"Graphical model","level":2,"score":0.32420000433921814},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3073999881744385},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.299699991941452},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.2978000044822693},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2890999913215637},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2793000042438507},{"id":"https://openalex.org/C124304363","wikidata":"https://www.wikidata.org/wiki/Q673661","display_name":"Abstraction","level":2,"score":0.27559998631477356},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.2606000006198883}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.26041","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.26041","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,"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":"doi:10.48550/arxiv.2603.26041","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.26041","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,"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"In":[0,41],"recent":[1,175],"years,":[2],"GUI":[3,22,56,75,127,158,206],"visual":[4,29,207],"agents":[5],"built":[6],"upon":[7],"Multimodal":[8],"Large":[9],"Language":[10],"Models":[11],"(MLLMs)":[12],"have":[13,113],"demonstrated":[14],"strong":[15],"potential":[16],"in":[17,55,142],"navigation":[18],"tasks.":[19],"However,":[20],"high-resolution":[21],"screenshots":[23,54,76,95,176],"produce":[24],"a":[25,78,89,161],"large":[26],"number":[27],"of":[28,35,204],"tokens,":[30],"making":[31],"the":[32,107,150,202],"direct":[33],"preservation":[34],"complete":[36],"historical":[37,53],"information":[38],"computationally":[39],"expensive.":[40],"this":[42,85],"paper,":[43],"we":[44,72,87,102,155,182],"conduct":[45],"an":[46,139],"empirical":[47],"study":[48],"on":[49],"token":[50,171],"pruning":[51,69,134,137],"for":[52,66,126,201],"scenarios":[57],"and":[58,98,177,198],"distill":[59],"three":[60],"practical":[61,199],"insights":[62,197],"that":[63,74,110,157],"are":[64],"crucial":[65],"designing":[67],"effective":[68],"strategies.":[70],"First,":[71],"observe":[73,156],"exhibit":[77,160],"distinctive":[79],"foreground-background":[80],"semantic":[81,115],"composition.":[82],"To":[83],"probe":[84],"property,":[86],"apply":[88],"simple":[90],"edge-based":[91],"separation":[92],"to":[93,106,165,173],"partition":[94],"into":[96],"foreground":[97],"background":[99,111],"regions.":[100],"Surprisingly,":[101],"find":[103],"that,":[104],"contrary":[105],"common":[108],"assumption":[109],"areas":[112],"little":[114],"value,":[116],"they":[117],"effectively":[118],"capture":[119],"interface-state":[120],"transitions,":[121],"thereby":[122],"providing":[123],"auxiliary":[124],"cues":[125],"reasoning.":[128],"Second,":[129],"compared":[130],"with":[131],"carefully":[132],"designed":[133],"strategies,":[135],"random":[136],"possesses":[138],"inherent":[140],"advantage":[141],"preserving":[143],"spatial":[144],"structure,":[145],"enabling":[146],"better":[147],"performance":[148],"under":[149],"same":[151],"computational":[152,186],"budget.":[153],"Finally,":[154],"Agents":[159],"recency":[162],"effect":[163],"similar":[164],"human":[166],"cognition:":[167],"by":[168],"allocating":[169],"larger":[170],"budgets":[172],"more":[174],"heavily":[178],"compressing":[179],"distant":[180],"ones,":[181],"can":[183],"significantly":[184],"reduce":[185],"cost":[187],"while":[188],"maintaining":[189],"nearly":[190],"unchanged":[191],"performance.":[192],"These":[193],"findings":[194],"offer":[195],"new":[196],"guidance":[200],"design":[203],"efficient":[205],"agents.":[208]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-03-31T00:00:00"}