{"id":"https://openalex.org/W7162780508","doi":"https://doi.org/10.48550/arxiv.2605.29438","title":"ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models","display_name":"ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models","publication_year":2026,"publication_date":"2026-05-28","ids":{"openalex":"https://openalex.org/W7162780508","doi":"https://doi.org/10.48550/arxiv.2605.29438"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.29438","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29438","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.2605.29438","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137341249","display_name":"Ye Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Ye","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137358834","display_name":"Huanan Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Huanan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137354382","display_name":"Kangye Ji","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji, Kangye","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137384371","display_name":"Yuan Meng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Meng, Yuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137372454","display_name":"Jiajun Fan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fan, Jiajun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123233473","display_name":"Yuansong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Yuansong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137377943","display_name":"Shiyu Qin","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qin, Shiyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5039066582","display_name":"Chenglei Wu","orcid":"https://orcid.org/0000-0002-7307-9480"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Chenglei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137391528","display_name":"Shu-Tao Xia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xia, Shu-Tao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137311758","display_name":"Zhi Wang","orcid":"https://orcid.org/0009-0004-9177-8172"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zhi","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":0.5496000051498413,"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":0.5496000051498413,"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/T10653","display_name":"Robot Manipulation and Learning","score":0.1890999972820282,"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/T10709","display_name":"Social Robot Interaction and HRI","score":0.038600001484155655,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5796999931335449},{"id":"https://openalex.org/keywords/representation","display_name":"Representation (politics)","score":0.5195000171661377},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.4733999967575073},{"id":"https://openalex.org/keywords/acceleration","display_name":"Acceleration","score":0.46070000529289246},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.42340001463890076},{"id":"https://openalex.org/keywords/action","display_name":"Action (physics)","score":0.4058000147342682},{"id":"https://openalex.org/keywords/reuse","display_name":"Reuse","score":0.37470000982284546},{"id":"https://openalex.org/keywords/robotics","display_name":"Robotics","score":0.3481000065803528},{"id":"https://openalex.org/keywords/theory-of-computation","display_name":"Theory of computation","score":0.3411000072956085}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7208999991416931},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5796999931335449},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.5195000171661377},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4950000047683716},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4733999967575073},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.46070000529289246},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.42340001463890076},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.4058000147342682},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3930000066757202},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.37470000982284546},{"id":"https://openalex.org/C34413123","wikidata":"https://www.wikidata.org/wiki/Q170978","display_name":"Robotics","level":3,"score":0.3481000065803528},{"id":"https://openalex.org/C24858836","wikidata":"https://www.wikidata.org/wiki/Q844718","display_name":"Theory of computation","level":2,"score":0.3411000072956085},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.32269999384880066},{"id":"https://openalex.org/C90509273","wikidata":"https://www.wikidata.org/wiki/Q11012","display_name":"Robot","level":2,"score":0.30959999561309814},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.3025999963283539},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.2996000051498413},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.2906000018119812},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.2858000099658966},{"id":"https://openalex.org/C104114177","wikidata":"https://www.wikidata.org/wiki/Q79782","display_name":"Motion (physics)","level":2,"score":0.28369998931884766},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.2702000141143799},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.26840001344680786},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.26510000228881836},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.25999999046325684},{"id":"https://openalex.org/C2778712577","wikidata":"https://www.wikidata.org/wiki/Q3505966","display_name":"Retraining","level":2,"score":0.25859999656677246},{"id":"https://openalex.org/C132459708","wikidata":"https://www.wikidata.org/wiki/Q744069","display_name":"Extrapolation","level":2,"score":0.25380000472068787},{"id":"https://openalex.org/C91575142","wikidata":"https://www.wikidata.org/wiki/Q1971426","display_name":"Optimal control","level":2,"score":0.25060001015663147}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.29438","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29438","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.2605.29438","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29438","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":{"Vision-Language-Action":[0],"(VLA)":[1],"models":[2,88,113],"are":[3],"a":[4,105,121,152,174,198],"powerful":[5],"paradigm":[6],"for":[7,189,202],"generalist":[8],"robotic":[9,22],"control.":[10,69],"However,":[11],"their":[12],"high":[13],"computational":[14],"cost":[15],"and":[16,29,60,77,96,131,143,220,226,229],"limited":[17],"control":[18,36,54,242],"frequency":[19,243],"hinder":[20],"real-time":[21],"manipulation,":[23],"especially":[24],"when":[25,91,97],"large":[26],"vision-language":[27],"backbones":[28],"iterative":[30],"action":[31,144,170],"heads":[32],"run":[33],"at":[34],"every":[35],"step.":[37],"Existing":[38],"VLA":[39,87,112,204],"acceleration":[40,50,200],"methods":[41],"often":[42],"optimize":[43],"individual":[44],"components":[45],"or":[46,212],"rely":[47],"on":[48,81,165,218,230],"fixed":[49,58],"rules,":[51],"treating":[52],"different":[53],"steps":[55],"with":[56,206],"largely":[57],"computation":[59,95,137,237],"overlooking":[61],"the":[62,139,149,214],"non-uniform":[63],"reasoning":[64],"demands":[65],"of":[66],"sequential":[67],"embodied":[68],"Inspired":[70],"by":[71,238],"human":[72],"motor":[73],"control,":[74],"where":[75],"cognitive":[76],"feedback":[78],"resources":[79],"concentrate":[80],"goal-sensitive":[82,190],"stages,":[83],"we":[84],"argue":[85],"that":[86,110,124],"should":[89],"learn":[90],"to":[92,98,134,160,224,247],"invest":[93],"full":[94,158,187],"reuse":[99],"prior":[100],"computation.":[101],"We":[102],"propose":[103],"ElegantVLA,":[104],"plug-in":[106],"phase-adaptive":[107],"inference":[108],"framework":[109,201],"accelerates":[111],"through":[114],"intra-model":[115],"dynamic":[116],"compute":[117,155],"scheduling.":[118],"ElegantVLA":[119,196,235],"introduces":[120],"lightweight":[122],"scheduler":[123,150],"observes":[125],"temporal":[126,162],"representation":[127,167],"similarity,":[128],"robot-motion":[129],"cues,":[130],"episode":[132],"progress":[133],"jointly":[135],"allocate":[136],"across":[138],"vision":[140],"encoder,":[141],"LLM,":[142],"head.":[145],"For":[146,169],"perception-language":[147],"reasoning,":[148],"selects":[151,173],"five-level":[153],"Vision-LLM":[154],"mode,":[156,177],"from":[157,244],"recomputation":[159],"multi-step":[161],"reuse,":[163],"based":[164],"visual-language":[166],"stability.":[168],"generation,":[171],"it":[172],"three-level":[175],"denoising":[176,180],"reusing":[178],"intermediate":[179],"states":[181],"during":[182],"stable":[183],"motion":[184],"while":[185,240],"preserving":[186],"refinement":[188],"stages.":[191],"By":[192],"coordinating":[193],"these":[194],"decisions,":[195],"offers":[197],"general":[199],"modern":[203],"pipelines":[205],"explicit":[207],"action-generation":[208],"modules,":[209],"without":[210],"modifying":[211],"retraining":[213],"base":[215],"model.":[216],"Experiments":[217],"GR00T":[219,233],"CogACT":[221],"achieve":[222],"up":[223],"2.55x":[225],"3.77x":[227],"speedup,":[228],"six":[231],"real-world":[232],"tasks":[234],"cuts":[236],"2.18x":[239],"raising":[241],"13.8":[245],"Hz":[246],"26.3":[248],"Hz.":[249]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-30T00:00:00"}