{"id":"https://openalex.org/W7160834051","doi":"https://doi.org/10.48550/arxiv.2605.07817","title":"GazeVLM: Active Vision via Internal Attention Control for Multimodal Reasoning","display_name":"GazeVLM: Active Vision via Internal Attention Control for Multimodal Reasoning","publication_year":2026,"publication_date":"2026-05-08","ids":{"openalex":"https://openalex.org/W7160834051","doi":"https://doi.org/10.48550/arxiv.2605.07817"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.07817","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07817","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","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.2605.07817","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135868376","display_name":"Brown Ebouky","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ebouky, Brown","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135878769","display_name":"Gabriele Carrino","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Carrino, Gabriele","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5105102978","display_name":"Niccolo Avogaro","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Avogaro, Niccolo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083617223","display_name":"Christoph Studer","orcid":"https://orcid.org/0000-0001-8950-6267"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Studer, Christoph","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017619464","display_name":"Andrea Bartezzaghi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bartezzaghi, Andrea","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5081069627","display_name":"Mattia Rigotti","orcid":"https://orcid.org/0000-0001-6466-2810"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rigotti, Mattia","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":0.7645000219345093,"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":0.7645000219345093,"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":0.10589999705553055,"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/T11707","display_name":"Gaze Tracking and Assistive Technology","score":0.05469999834895134,"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/active-vision","display_name":"Active vision","score":0.6280999779701233},{"id":"https://openalex.org/keywords/visual-reasoning","display_name":"Visual reasoning","score":0.5353999733924866},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5335000157356262},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.5097000002861023},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.46720001101493835},{"id":"https://openalex.org/keywords/focus","display_name":"Focus (optics)","score":0.460099995136261},{"id":"https://openalex.org/keywords/gaze","display_name":"Gaze","score":0.4406999945640564},{"id":"https://openalex.org/keywords/spatial-contextual-awareness","display_name":"Spatial contextual awareness","score":0.42080000042915344},{"id":"https://openalex.org/keywords/visual-attention","display_name":"Visual attention","score":0.41850000619888306},{"id":"https://openalex.org/keywords/control","display_name":"Control (management)","score":0.4133000075817108}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7064999938011169},{"id":"https://openalex.org/C193611912","wikidata":"https://www.wikidata.org/wiki/Q4677596","display_name":"Active vision","level":2,"score":0.6280999779701233},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.5353999733924866},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5335000157356262},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.5097000002861023},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.5023999810218811},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.46720001101493835},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.460099995136261},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.45809999108314514},{"id":"https://openalex.org/C2779916870","wikidata":"https://www.wikidata.org/wiki/Q14467155","display_name":"Gaze","level":2,"score":0.4406999945640564},{"id":"https://openalex.org/C64754055","wikidata":"https://www.wikidata.org/wiki/Q7574053","display_name":"Spatial contextual awareness","level":2,"score":0.42080000042915344},{"id":"https://openalex.org/C2986089797","wikidata":"https://www.wikidata.org/wiki/Q6501338","display_name":"Visual attention","level":3,"score":0.41850000619888306},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.4133000075817108},{"id":"https://openalex.org/C155911833","wikidata":"https://www.wikidata.org/wiki/Q3817354","display_name":"Spatial intelligence","level":2,"score":0.3806000053882599},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3659000098705292},{"id":"https://openalex.org/C56461940","wikidata":"https://www.wikidata.org/wiki/Q970687","display_name":"Eye tracking","level":2,"score":0.35929998755455017},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.35690000653266907},{"id":"https://openalex.org/C44210515","wikidata":"https://www.wikidata.org/wiki/Q16968978","display_name":"Bespoke","level":2,"score":0.3321000039577484},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.33059999346733093},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3098999857902527},{"id":"https://openalex.org/C90509273","wikidata":"https://www.wikidata.org/wiki/Q11012","display_name":"Robot","level":2,"score":0.3050999939441681},{"id":"https://openalex.org/C207347870","wikidata":"https://www.wikidata.org/wiki/Q371174","display_name":"Gesture","level":2,"score":0.29030001163482666},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.28600001335144043},{"id":"https://openalex.org/C97364631","wikidata":"https://www.wikidata.org/wiki/Q484284","display_name":"Deductive reasoning","level":2,"score":0.28380000591278076},{"id":"https://openalex.org/C178253425","wikidata":"https://www.wikidata.org/wiki/Q162668","display_name":"Visual perception","level":3,"score":0.28369998931884766},{"id":"https://openalex.org/C194995250","wikidata":"https://www.wikidata.org/wiki/Q531136","display_name":"Affordance","level":2,"score":0.2825999855995178},{"id":"https://openalex.org/C126139899","wikidata":"https://www.wikidata.org/wiki/Q4818236","display_name":"Attentional control","level":3,"score":0.27799999713897705},{"id":"https://openalex.org/C13077596","wikidata":"https://www.wikidata.org/wiki/Q340015","display_name":"Deixis","level":2,"score":0.27790001034736633},{"id":"https://openalex.org/C30181142","wikidata":"https://www.wikidata.org/wiki/Q865103","display_name":"Foveal","level":3,"score":0.2745000123977661},{"id":"https://openalex.org/C115086926","wikidata":"https://www.wikidata.org/wiki/Q17004651","display_name":"Causal reasoning","level":3,"score":0.2718999981880188},{"id":"https://openalex.org/C101468663","wikidata":"https://www.wikidata.org/wiki/Q1620158","display_name":"Modular design","level":2,"score":0.26820001006126404},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.26759999990463257},{"id":"https://openalex.org/C118147538","wikidata":"https://www.wikidata.org/wiki/Q1126970","display_name":"Metacognition","level":3,"score":0.26350000500679016},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.26159998774528503},{"id":"https://openalex.org/C104114177","wikidata":"https://www.wikidata.org/wiki/Q79782","display_name":"Motion (physics)","level":2,"score":0.25679999589920044},{"id":"https://openalex.org/C185761835","wikidata":"https://www.wikidata.org/wiki/Q1431287","display_name":"Peripheral vision","level":2,"score":0.2542000114917755}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.07817","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07817","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.07817","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07817","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","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":{"Human":[0],"visual":[1,39,123,177,182],"reasoning":[2,54,84,135,160,205],"is":[3],"governed":[4],"by":[5,214,226],"active":[6],"vision,":[7],"a":[8,67,98,116,186],"process":[9,38],"where":[10],"metacognitive":[11,73],"control":[12,100],"drives":[13],"top-down":[14,99],"goal-directed":[15],"attention,":[16],"dynamically":[17,110],"routing":[18],"foveal":[19,131],"focus":[20],"toward":[21],"task-relevant":[22],"details":[23],"while":[24],"maintaining":[25],"peripheral":[26],"awareness":[27,156],"of":[28,47,78],"the":[29,44,62,83,88,137,142,148,172],"global":[30,143,154],"scene.":[31],"In":[32],"contrast,":[33],"modern":[34],"Vision-Language":[35],"Models":[36],"(VLMs)":[37],"information":[40],"passively,":[41],"relying":[42,162],"on":[43,163,230],"static":[45],"accumulation":[46],"massive":[48],"token":[49],"contexts":[50],"that":[51,70,120,194],"dilute":[52],"spatial":[53,126,155],"and":[55,129,157,217,232],"induce":[56],"linguistic":[57],"hallucinations.":[58],"Here":[59],"we":[60],"propose":[61],"following":[63],"paradigm":[64],"shift:":[65],"GazeVLM,":[66],"multimodal":[68,204,219],"architecture":[69,146],"internalizes":[71],"this":[72],"oversight":[74],"over":[75,102],"its":[76,103,112,211],"deployment":[77],"attention":[79,106,128],"resources":[80],"directly":[81],"into":[82],"loop.":[85],"By":[86],"empowering":[87],"VLM":[89],"to":[90,150],"autonomously":[91],"generate":[92],"gaze":[93],"tokens":[94,178],"($\\texttt{}$),":[95],"GazeVLM":[96,200],"establishes":[97],"mechanism":[101],"own":[104],"causal":[105],"mask.":[107],"The":[108],"model":[109,149],"dictates":[111],"focal":[113,159],"intent,":[114],"triggering":[115],"continuous":[117],"suppression":[118],"bias":[119,138],"dampens":[121],"irrelevant":[122],"features,":[124],"implementing":[125],"selective":[127],"simulating":[130],"fixation.":[132],"Once":[133],"local":[134],"concludes,":[136],"lifts,":[139],"seamlessly":[140],"restoring":[141],"view.":[144],"This":[145],"enables":[147],"fluidly":[151],"transition":[152],"between":[153],"localized":[158,181],"without":[161],"external":[164],"agentic":[165,218],"contraptions":[166],"like":[167],"cropping":[168],"tools,":[169],"or":[170],"inflating":[171],"context":[173],"window":[174],"with":[175,185,224],"additional":[176],"derived":[179],"from":[180],"patches.":[183],"Trained":[184],"bespoke":[187],"Group":[188],"Relative":[189],"Policy":[190],"Optimization":[191],"(GRPO)":[192],"procedure":[193],"rewards":[195],"valid":[196],"grounding,":[197],"our":[198],"4B-parameter":[199],"delivers":[201],"strong":[202],"high-resolution":[203],"performance,":[206],"surpassing":[207],"state-of-the-art":[208],"VLMs":[209],"in":[210],"parameter":[212],"class":[213],"nearly":[215],"4%":[216],"pipelines":[220],"built":[221],"around":[222],"thinking":[223],"images":[225],"more":[227],"than":[228],"5%":[229],"HRBench-4k":[231],"HRBench-8k.":[233]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-12T00:00:00"}