{"id":"https://openalex.org/W7204671285","doi":"https://doi.org/10.48550/arxiv.2608.27384","title":"FlashVLA: Streaming Action Decoding for Fast and Asynchronous VLA Inference","display_name":"FlashVLA: Streaming Action Decoding for Fast and Asynchronous VLA Inference","alternate_titles":[],"publication_year":2026,"publication_date":"2026-08-27","ids":{"openalex":"https://openalex.org/W7204671285","doi":"https://doi.org/10.48550/arxiv.2608.27384"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.27384","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.27384","pdf_url":"https://arxiv.org/pdf/2608.27384","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"parent_works":[],"lineage":["https://openalex.org/W7204671285"],"child_works_count":0,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.27384","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5148603579","display_name":"Zekai Li","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Zekai","raw_affiliation_strings":["MIT https ://github.com/ - lab/flashvla.git"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"MIT https ://github.com/ - lab/flashvla.git","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103244882","display_name":"Jiaming Tang","orcid":"https://orcid.org/0009-0004-4186-6561","observed_orcids":["https://orcid.org/0009-0004-4186-6561"]},"institutions":[{"id":"https://openalex.org/I36258959","display_name":"University of California San Diego","ror":"https://ror.org/0168r3w48","country_code":"US","type":"education","lineage":["https://openalex.org/I36258959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Tang, Jiaming","raw_affiliation_strings":["UC San Diego"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"UC San Diego","institution_ids":["https://openalex.org/I36258959"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100750264","display_name":"Zhijian Liu","orcid":"https://orcid.org/0000-0003-3632-9986","observed_orcids":["https://orcid.org/0000-0003-3632-9986"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Zhijian","raw_affiliation_strings":["MIT https ://github.com/ - lab/flashvla.git"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"MIT https ://github.com/ - lab/flashvla.git","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":9,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_pulled":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10653","display_name":"Robot Manipulation and Learning","score":0.9330999851226807,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10653","display_name":"Robot Manipulation and Learning","score":0.9330999851226807,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.02930000051856041,"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"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.014700000174343586,"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/vision-language-action-models","display_name":"vision-language-action models","score":0.9819999933242798},{"id":"https://openalex.org/keywords/robotic-manipulation","display_name":"robotic manipulation","score":0.7509999871253967},{"id":"https://openalex.org/keywords/asynchronous-inference","display_name":"asynchronous inference","score":0.8339999914169312},{"id":"https://openalex.org/keywords/low-latency-inference","display_name":"low-latency inference","score":0.8190000057220459},{"id":"https://openalex.org/keywords/flow-matching","display_name":"flow matching","score":0.7639999985694885},{"id":"https://openalex.org/keywords/high-frequency-control","display_name":"high-frequency control","score":0.45500001311302185},{"id":"https://openalex.org/keywords/vision-language-models","display_name":"vision-language models","score":0.5}],"concepts":[{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.90420001745224},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8217999935150146},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7592999935150146},{"id":"https://openalex.org/C160145156","wikidata":"https://www.wikidata.org/wiki/Q778586","display_name":"Executable","level":2,"score":0.626800000667572},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.6216999888420105},{"id":"https://openalex.org/C2778858076","wikidata":"https://www.wikidata.org/wiki/Q5249539","display_name":"Decodes","level":3,"score":0.4772000014781952},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.42910000681877136},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.4059999883174896},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.336899995803833},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.30709999799728394},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.3003999888896942},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.29840001463890076},{"id":"https://openalex.org/C7923308","wikidata":"https://www.wikidata.org/wiki/Q4812211","display_name":"Asynchronous system","level":5,"score":0.29260000586509705},{"id":"https://openalex.org/C75291252","wikidata":"https://www.wikidata.org/wiki/Q1315756","display_name":"TRACE (psycholinguistics)","level":2,"score":0.28380000591278076},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.28040000796318054},{"id":"https://openalex.org/C134261354","wikidata":"https://www.wikidata.org/wiki/Q938438","display_name":"Statistical inference","level":2,"score":0.2750999927520752},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.2669999897480011},{"id":"https://openalex.org/C134652429","wikidata":"https://www.wikidata.org/wiki/Q1052698","display_name":"Jitter","level":2,"score":0.2597000002861023},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.25699999928474426},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.25519999861717224}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.27384","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.27384","pdf_url":"https://arxiv.org/pdf/2608.27384","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.27384","doi":"https://doi.org/10.48550/arxiv.2608.27384","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.27384","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,"listed_in":[],"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.27384","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.27384","pdf_url":"https://arxiv.org/pdf/2608.27384","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204671285.pdf","grobid_xml":"https://content.openalex.org/works/W7204671285.grobid-xml"},"referenced_works_count":17,"referenced_works":["https://openalex.org/W2734498959","https://openalex.org/W4399695759","https://openalex.org/W4407874080","https://openalex.org/W4414674749","https://openalex.org/W4414898213","https://openalex.org/W4415163123","https://openalex.org/W4416958104","https://openalex.org/W4417097926","https://openalex.org/W4417144613","https://openalex.org/W4417287217","https://openalex.org/W6929480010","https://openalex.org/W7103171635","https://openalex.org/W7125832496","https://openalex.org/W7127648241","https://openalex.org/W7133220882","https://openalex.org/W7139920986","https://openalex.org/W7147084562"],"related_works":["https://openalex.org/W7214746916","https://openalex.org/W4416958104","https://openalex.org/W7147084562","https://openalex.org/W7160963588","https://openalex.org/W4415037022","https://openalex.org/W7117248983","https://openalex.org/W7168380537","https://openalex.org/W7164917509","https://openalex.org/W7212656183","https://openalex.org/W7167680094"],"abstract_inverted_index":{"Vision-Language-Action":[0],"(VLA)":[1],"models":[2],"are":[3],"increasingly":[4],"promising":[5],"for":[6],"robotic":[7],"manipulation,":[8],"yet":[9],"their":[10],"real-world":[11,143,170],"deployment":[12],"remains":[13],"bottlenecked":[14],"by":[15],"high":[16],"inference":[17,46,66,120,148,168],"latency":[18],"and":[19,51,67,102,142],"unstable":[20],"asynchronous":[21,52,71,133,167],"execution.":[22,72],"This":[23,109],"challenge":[24],"is":[25],"particularly":[26],"pronounced":[27],"in":[28,85,169],"flow-matching-based":[29],"VLA":[30],"models,":[31],"where":[32],"action":[33,78,93,117,129],"decoding":[34,79],"requires":[35],"multiple":[36,96],"iterative":[37],"steps":[38],"conditioned":[39],"on":[40,161],"the":[41],"VLM":[42],"context.":[43],"While":[44],"efficient":[45],"methods":[47,53],"improve":[48],"control":[49,159],"frequency":[50,160],"reduce":[54],"execution":[55,134],"idle":[56],"time,":[57],"existing":[58],"approaches":[59],"often":[60],"fail":[61],"to":[62,113],"jointly":[63],"achieve":[64,157],"low-latency":[65],"accurate,":[68],"temporally":[69],"consistent":[70],"We":[73],"introduce":[74],"\\textbf{FlashVLA},":[75],"a":[76,86,91,162],"streaming":[77,92],"framework":[80],"that":[81],"addresses":[82],"both":[83],"challenges":[84],"unified":[87],"formulation.":[88],"FlashVLA":[89,112,145],"maintains":[90],"buffer":[94],"with":[95,165],"chunks":[97],"at":[98],"different":[99],"noise":[100],"levels":[101],"decodes":[103],"them":[104],"using":[105],"chunk-wise":[106,124],"causal":[107],"attention.":[108],"design":[110],"allows":[111],"produce":[114],"one":[115],"executable":[116],"chunk":[118],"per":[119],"step.":[121],"Moreover,":[122],"its":[123],"autoregressive":[125],"formulation":[126],"implicitly":[127],"preserves":[128],"continuity,":[130],"enabling":[131],"smooth":[132,166],"without":[135],"extra":[136],"future-state":[137],"conditioning.":[138],"Across":[139],"extensive":[140],"simulated":[141],"experiments,":[144],"substantially":[146],"improves":[147],"speed":[149],"while":[150],"maintaining":[151],"strong":[152],"task":[153],"performance.":[154],"It":[155],"can":[156],"$\\geq$30\\,Hz":[158],"single":[163],"GPU":[164],"deployment.":[171]},"cited_by_api_url":"https://api.openalex.org/works?filter=cites:W7204671285","counts_by_year":[{"year":2026,"cited_by_count":9}],"updated_date":"2026-11-17T12:49:18.490984","created_date":"2026-08-30T00:00:00"}
