{"id":"https://openalex.org/W7154016775","doi":"https://doi.org/10.48550/arxiv.2604.09508","title":"VISOR: Agentic Visual Retrieval-Augmented Generation via Iterative Search and Over-horizon Reasoning","display_name":"VISOR: Agentic Visual Retrieval-Augmented Generation via Iterative Search and Over-horizon Reasoning","publication_year":2026,"publication_date":"2026-04-10","ids":{"openalex":"https://openalex.org/W7154016775","doi":"https://doi.org/10.48550/arxiv.2604.09508"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.09508","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.09508","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":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.2604.09508","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5085330182","display_name":"Yi Shen","orcid":"https://orcid.org/0000-0001-6539-4918"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shen, Yucheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125970112","display_name":"Jiulong Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Jiulong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133476229","display_name":"Jizhou Huang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Jizhou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133479895","display_name":"Dawei Yin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yin, Dawei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133495851","display_name":"Lingyong Yan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yan, Lingyong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133501831","display_name":"Min Cao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cao, Min","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":0.9919000267982483,"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":0.9919000267982483,"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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.0031999999191612005,"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":0.0007999999797903001,"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-search","display_name":"Visual search","score":0.7419000267982483},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.6057999730110168},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.5843999981880188},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.48730000853538513},{"id":"https://openalex.org/keywords/trajectory","display_name":"Trajectory","score":0.444599986076355},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.3806000053882599},{"id":"https://openalex.org/keywords/iterative-refinement","display_name":"Iterative refinement","score":0.37549999356269836},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.36809998750686646},{"id":"https://openalex.org/keywords/action","display_name":"Action (physics)","score":0.3659999966621399},{"id":"https://openalex.org/keywords/masking","display_name":"Masking (illustration)","score":0.3605000078678131}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8091999888420105},{"id":"https://openalex.org/C158495155","wikidata":"https://www.wikidata.org/wiki/Q2369151","display_name":"Visual search","level":2,"score":0.7419000267982483},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.6057999730110168},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.5843999981880188},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5432999730110168},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.48730000853538513},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.444599986076355},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3806000053882599},{"id":"https://openalex.org/C2779982483","wikidata":"https://www.wikidata.org/wiki/Q6094420","display_name":"Iterative refinement","level":2,"score":0.37549999356269836},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.36809998750686646},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.3659999966621399},{"id":"https://openalex.org/C2777402240","wikidata":"https://www.wikidata.org/wiki/Q6783436","display_name":"Masking (illustration)","level":2,"score":0.3605000078678131},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.35989999771118164},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3504999876022339},{"id":"https://openalex.org/C169900460","wikidata":"https://www.wikidata.org/wiki/Q2200417","display_name":"Cognition","level":2,"score":0.3490999937057495},{"id":"https://openalex.org/C207363949","wikidata":"https://www.wikidata.org/wiki/Q462915","display_name":"Visual space","level":3,"score":0.34610000252723694},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.3458000123500824},{"id":"https://openalex.org/C87833898","wikidata":"https://www.wikidata.org/wiki/Q1060280","display_name":"Advanced driver assistance systems","level":2,"score":0.34290000796318054},{"id":"https://openalex.org/C56461940","wikidata":"https://www.wikidata.org/wiki/Q970687","display_name":"Eye tracking","level":2,"score":0.33719998598098755},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3165999948978424},{"id":"https://openalex.org/C72434380","wikidata":"https://www.wikidata.org/wiki/Q230930","display_name":"State space","level":2,"score":0.31349998712539673},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.31209999322891235},{"id":"https://openalex.org/C164280684","wikidata":"https://www.wikidata.org/wiki/Q5529040","display_name":"Gaze-contingency paradigm","level":4,"score":0.3084000051021576},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.2946999967098236},{"id":"https://openalex.org/C178253425","wikidata":"https://www.wikidata.org/wiki/Q162668","display_name":"Visual perception","level":3,"score":0.28760001063346863},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.28139999508857727},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.2800999879837036},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2759999930858612},{"id":"https://openalex.org/C125583679","wikidata":"https://www.wikidata.org/wiki/Q755673","display_name":"Search algorithm","level":2,"score":0.2709999978542328},{"id":"https://openalex.org/C64876066","wikidata":"https://www.wikidata.org/wiki/Q5141226","display_name":"Cognitive neuroscience of visual object recognition","level":3,"score":0.2637999951839447},{"id":"https://openalex.org/C112313634","wikidata":"https://www.wikidata.org/wiki/Q7886648","display_name":"Complement (music)","level":5,"score":0.2619999945163727},{"id":"https://openalex.org/C64754055","wikidata":"https://www.wikidata.org/wiki/Q7574053","display_name":"Spatial contextual awareness","level":2,"score":0.25920000672340393},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.25699999928474426},{"id":"https://openalex.org/C143587482","wikidata":"https://www.wikidata.org/wiki/Q1543216","display_name":"Iterative and incremental development","level":2,"score":0.25119999051094055}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.09508","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.09508","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":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.2604.09508","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.09508","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"display_name":"Peace, Justice and strong institutions","id":"https://metadata.un.org/sdg/16","score":0.498752623796463}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Visual":[0,39,130],"Retrieval-Augmented":[1,105],"Generation":[2,106],"(VRAG)":[3],"empowers":[4],"Vision-Language":[5],"Models":[6],"to":[7,91,136,152],"retrieve":[8],"and":[9,85,110,133,149,191,204],"reason":[10],"over":[11],"visually":[12],"rich":[13],"documents.":[14],"To":[15,97],"tackle":[16],"complex":[17],"queries":[18],"requiring":[19],"multi-step":[20],"reasoning,":[21,126],"agentic":[22,32],"VRAG":[23,33],"systems":[24],"interleave":[25],"reasoning":[26,218],"with":[27,128,146,188,212],"iterative":[28],"retrieval..":[29],"However,":[30],"existing":[31],"faces":[34],"two":[35],"critical":[36],"bottlenecks.":[37],"(1)":[38],"Evidence":[40,121],"Sparsity:":[41],"key":[42],"evidence":[43,58,159],"is":[44],"scattered":[45],"across":[46,80],"pages":[47,82],"yet":[48],"processed":[49],"in":[50,72],"isolation,":[51],"hindering":[52],"cross-page":[53,125],"reasoning;":[54],"moreover,":[55],"fine-grained":[56],"intra-image":[57],"often":[59],"requires":[60],"precise":[61],"visual":[62,78,138,172,217],"actions,":[63],"whose":[64],"misuse":[65],"degrades":[66],"retrieval":[67],"quality;":[68],"(2)":[69],"Search":[70,109],"Drift":[71],"Long":[73],"Horizons:":[74],"the":[75,158],"accumulation":[76],"of":[77],"tokens":[79],"retrieved":[81],"dilutes":[83],"context":[84,167,197],"causes":[86],"cognitive":[87],"overload,":[88],"leading":[89],"agents":[90],"deviate":[92],"from":[93,168],"their":[94],"search":[95,154],"objective.":[96],"address":[98],"these":[99],"challenges,":[100],"we":[101,141],"propose":[102],"VISOR":[103,117,176,208],"(Visual":[104],"via":[107],"Iterative":[108],"Over-horizon":[111],"Reasoning),":[112],"a":[113,119,129,143,178],"unified":[114],"single-agent":[115],"framework.":[116],"features":[118],"structured":[120],"Space":[122],"for":[123,195,215],"progressive":[124],"coupled":[127],"Action":[131],"Evaluation":[132],"Correction":[134],"mechanism":[135],"manage":[137],"actions.":[139],"Additionally,":[140],"introduce":[142],"Dynamic":[144],"Trajectory":[145],"Sliding":[147],"Window":[148],"Intent":[150],"Injection":[151],"mitigate":[153],"drift.":[155],"They":[156],"anchor":[157],"space":[160],"while":[161],"discarding":[162],"earlier":[163],"raw":[164],"interactions,":[165],"preventing":[166],"being":[169],"overwhelmed":[170],"by":[171],"tokens.":[173],"We":[174],"train":[175],"using":[177],"Group":[179],"Relative":[180],"Policy":[181],"Optimization-based":[182],"Reinforcement":[183],"Learning":[184],"(GRPO-based":[185],"RL)":[186],"pipeline":[187],"state":[189],"masking":[190],"credit":[192],"assignment":[193],"tailored":[194],"dynamic":[196],"reconstruction.":[198],"Extensive":[199],"experiments":[200],"on":[201],"ViDoSeek,":[202],"SlideVQA,":[203],"MMLongBench":[205],"demonstrate":[206],"that":[207],"achieves":[209],"state-of-the-art":[210],"performance":[211],"superior":[213],"efficiency":[214],"long-horizon":[216],"tasks.":[219]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-04-14T00:00:00"}