{"id":"https://openalex.org/W7162761625","doi":"https://doi.org/10.48550/arxiv.2605.30011","title":"VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies","display_name":"VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies","publication_year":2026,"publication_date":"2026-05-28","ids":{"openalex":"https://openalex.org/W7162761625","doi":"https://doi.org/10.48550/arxiv.2605.30011"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.30011","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30011","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.2605.30011","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5126413938","display_name":"Mingjian Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Mingjian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137347836","display_name":"Wenqiao Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Wenqiao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137328434","display_name":"Yuqian Yuan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuan, Yuqian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137340083","display_name":"Yang Dai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dai, Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137375189","display_name":"Binhe Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Binhe","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137385956","display_name":"Zheqi Lv","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lv, Zheqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137380589","display_name":"Haoyu Zheng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Haoyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137399694","display_name":"\u795d\u4f73\u797a ZHU Jiaqi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Jiaqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066775492","display_name":"Zhiqi Ge","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ge, Zhiqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137315842","display_name":"Zixuan Wan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wan, Zixuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137354059","display_name":"Siliang Tang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Siliang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137349454","display_name":"Yueting Zhuang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhuang, Yueting","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":0.9639999866485596,"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":0.9639999866485596,"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":0.0066999997943639755,"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":0.0052999998442828655,"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/counterfactual-thinking","display_name":"Counterfactual thinking","score":0.8046000003814697},{"id":"https://openalex.org/keywords/defeasible-estate","display_name":"Defeasible estate","score":0.7164999842643738},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.597100019454956},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.58160001039505},{"id":"https://openalex.org/keywords/bootstrapping","display_name":"Bootstrapping (finance)","score":0.5612000226974487},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.5523999929428101},{"id":"https://openalex.org/keywords/action","display_name":"Action (physics)","score":0.44859999418258667},{"id":"https://openalex.org/keywords/visualization","display_name":"Visualization","score":0.4124999940395355}],"concepts":[{"id":"https://openalex.org/C108650721","wikidata":"https://www.wikidata.org/wiki/Q1783253","display_name":"Counterfactual thinking","level":2,"score":0.8046000003814697},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7985000014305115},{"id":"https://openalex.org/C193856179","wikidata":"https://www.wikidata.org/wiki/Q5251100","display_name":"Defeasible estate","level":2,"score":0.7164999842643738},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.597100019454956},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.58160001039505},{"id":"https://openalex.org/C207609745","wikidata":"https://www.wikidata.org/wiki/Q4944086","display_name":"Bootstrapping (finance)","level":2,"score":0.5612000226974487},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.5523999929428101},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.44859999418258667},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4474000036716461},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.4124999940395355},{"id":"https://openalex.org/C100609095","wikidata":"https://www.wikidata.org/wiki/Q1335050","display_name":"Embodied cognition","level":2,"score":0.41040000319480896},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3917999863624573},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.3418000042438507},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.3158999979496002},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.31369999051094055},{"id":"https://openalex.org/C113843644","wikidata":"https://www.wikidata.org/wiki/Q901882","display_name":"Interface (matter)","level":4,"score":0.3125999867916107},{"id":"https://openalex.org/C2781020372","wikidata":"https://www.wikidata.org/wiki/Q533093","display_name":"On the fly","level":2,"score":0.3041999936103821},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.29600000381469727},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.295199990272522},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2736000120639801},{"id":"https://openalex.org/C46637626","wikidata":"https://www.wikidata.org/wiki/Q6693015","display_name":"Low latency (capital markets)","level":2,"score":0.25589999556541443},{"id":"https://openalex.org/C159877910","wikidata":"https://www.wikidata.org/wiki/Q2202883","display_name":"Autoregressive model","level":2,"score":0.2549000084400177}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.30011","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30011","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.2605.30011","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.30011","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"work":[1],"has":[2],"begun":[3],"to":[4,61,86,99,160,177],"equip":[5],"vision-language-action":[6],"(VLA)":[7],"policies":[8],"with":[9,30,64,175],"explicit":[10],"intermediate":[11],"reasoning.":[12],"In":[13],"embodied":[14],"control,":[15],"however,":[16],"textual":[17,26],"chain-of-thought":[18],"is":[19,60],"a":[20,48,73,94,116,121,125,180],"poor":[21],"fit:":[22],"irrelevant":[23],"or":[24],"weakly":[25],"information":[27],"can":[28],"interfere":[29],"action":[31,63,70],"prediction,":[32],"while":[33,81,108,152],"autoregressive":[34],"text":[35],"decoding":[36,83],"adds":[37],"too":[38],"much":[39],"latency":[40,156,172],"for":[41,52,130],"real-time":[42],"closed-loop":[43],"execution.":[44],"We":[45,112],"present":[46],"VISUALTHINK-VLA,":[47],"visual":[49,66,102],"intermediate-reasoning":[50],"framework":[51],"accurate,":[53],"low-latency":[54,106],"VLA":[55,127],"policies.":[56],"Our":[57],"bootstrapping":[58],"philosophy":[59],"guide":[62],"effective":[65],"thinking:":[67],"VISUALTHINK-VLA":[68,92,143],"bootstraps":[69],"prediction":[71],"through":[72],"compact":[74],"visual-evidence":[75],"interface":[76],"that":[77,123],"preserves":[78],"spatial":[79],"precision":[80],"avoiding":[82],"overhead.":[84],"Besides,":[85],"further":[87],"improve":[88],"performance":[89],"and":[90,133,140],"efficiency,":[91],"adopts":[93],"tailored":[95],"selective":[96],"routing":[97],"mechanism":[98],"learn":[100],"the":[101,145,154,161],"evidence":[103],"tokens,":[104],"enabling":[105],"inference":[107],"preserving":[109],"high-capacity":[110],"specialization.":[111],"also":[113],"introduce":[114],"VisualEvidence-Kit,":[115],"supervision-and-audit":[117],"resource":[118],"centered":[119],"on":[120,149,166],"VisualEvidence-Agent":[122],"constructs":[124],"754.7k":[126],"instructions":[128],"VisualEvidence-Set":[129],"route":[131],"supervision":[132],"counterfactual":[134],"faithfulness":[135],"tests.":[136],"Across":[137],"multiple":[138],"benchmarks":[139,151],"real-robot":[141],"evaluation,":[142],"achieves":[144],"highest":[146],"success":[147],"rate":[148],"most":[150],"reducing":[153],"multi-second":[155],"of":[157],"reasoning-augmented":[158],"baselines":[159],"sub-second":[162],"regime.":[163],"For":[164],"example,":[165],"BridgeData":[167],"V2,":[168],"it":[169],"reduces":[170],"step":[171],"from":[173],"8.377,s":[174],"ECoT":[176],"0.367,s,":[178],"achieving":[179],"22.8":[181],"times":[182],"speedup.":[183]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-30T00:00:00"}