{"id":"https://openalex.org/W7224768713","doi":"https://doi.org/10.48550/arxiv.2610.38508","title":"SaiVLA-JEPA: Evaluating Future Latent Prediction in JEPA-Conditioned VLA Control","display_name":"SaiVLA-JEPA: Evaluating Future Latent Prediction in JEPA-Conditioned VLA Control","publication_year":2026,"publication_date":"2026-10-28","ids":{"openalex":"https://openalex.org/W7224768713","doi":"https://doi.org/10.48550/arxiv.2610.38508"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.38508","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.38508","pdf_url":"https://arxiv.org/pdf/2610.38508","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":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2610.38508","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101220620","display_name":"Rufeng Xiao","orcid":"https://orcid.org/0009-0001-9474-7767","observed_orcids":["https://orcid.org/0009-0001-9474-7767"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiao, Rufeng","raw_affiliation_strings":["VLA Research Team, Synthoid.ai"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"VLA Research Team, Synthoid.ai","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017440452","display_name":"Ruihao Xiao","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiao, Ruihao","raw_affiliation_strings":["VLA Research Team, Synthoid.ai"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"VLA Research Team, Synthoid.ai","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5160229702","display_name":"Wei Chen","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Wei","raw_affiliation_strings":["VLA Research Team, Synthoid.ai"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"VLA Research Team, Synthoid.ai","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5089963091","display_name":"Xinhai Sun","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Xinhai","raw_affiliation_strings":["VLA Research Team, Synthoid.ai"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"VLA Research Team, Synthoid.ai","institution_ids":[]}]}],"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":true,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.54830002784729,"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"}},"topics":[{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.54830002784729,"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/T10653","display_name":"Robot Manipulation and Learning","score":0.35030001401901245,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.048500001430511475,"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/jepa","display_name":"JEPA","score":0.6320000290870667},{"id":"https://openalex.org/keywords/feedback-control","display_name":"feedback control","score":0.546999990940094},{"id":"https://openalex.org/keywords/parameter-adaptation","display_name":"parameter adaptation","score":0.8169999718666077}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.661300003528595},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4966000020503998},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47760000824928284},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.460099995136261},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.4246000051498413},{"id":"https://openalex.org/C90329073","wikidata":"https://www.wikidata.org/wiki/Q914232","display_name":"Ask price","level":2,"score":0.40849998593330383},{"id":"https://openalex.org/C172205157","wikidata":"https://www.wikidata.org/wiki/Q1782962","display_name":"Model predictive control","level":3,"score":0.29179999232292175},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.2831999957561493},{"id":"https://openalex.org/C45804977","wikidata":"https://www.wikidata.org/wiki/Q7239673","display_name":"Predictive modelling","level":2,"score":0.27799999713897705},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2773999869823456},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.2703999876976013},{"id":"https://openalex.org/C4679612","wikidata":"https://www.wikidata.org/wiki/Q866298","display_name":"Aggregate (composite)","level":2,"score":0.26919999718666077}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.38508","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.38508","pdf_url":"https://arxiv.org/pdf/2610.38508","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":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2610.38508","doi":"https://doi.org/10.48550/arxiv.2610.38508","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.38508","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-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2610.38508","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.38508","pdf_url":"https://arxiv.org/pdf/2610.38508","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":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7224768713.pdf","grobid_xml":"https://content.openalex.org/works/W7224768713.grobid-xml"},"referenced_works_count":17,"referenced_works":["https://openalex.org/W2084413241","https://openalex.org/W2795843265","https://openalex.org/W4385403811","https://openalex.org/W4385431288","https://openalex.org/W4386076428","https://openalex.org/W4390872297","https://openalex.org/W4402353985","https://openalex.org/W4409147637","https://openalex.org/W4414079054","https://openalex.org/W4417261359","https://openalex.org/W7133188016","https://openalex.org/W7133220882","https://openalex.org/W7137837393","https://openalex.org/W7167227348","https://openalex.org/W7201875857","https://openalex.org/W7202137576","https://openalex.org/W7202156618"],"related_works":["https://openalex.org/W7221015601","https://openalex.org/W7214310094","https://openalex.org/W7128621118","https://openalex.org/W7221089725","https://openalex.org/W7128802879","https://openalex.org/W7214936945","https://openalex.org/W7224090296","https://openalex.org/W7202156618","https://openalex.org/W7160875620","https://openalex.org/W7221541595"],"abstract_inverted_index":{"Joint-embedding":[0],"predictive":[1,25],"architectures":[2],"(JEPAs)":[3],"learn":[4],"to":[5,15],"predict":[6,70],"representations":[7,72],"rather":[8,178],"than":[9,179],"pixels,":[10],"offering":[11],"a":[12,31,49,53,62,147,151],"compact":[13],"way":[14],"model":[16,51],"how":[17],"visual":[18],"scenes":[19],"evolve.":[20],"We":[21,39,85,129],"ask":[22],"whether":[23],"this":[24],"objective":[26],"improves":[27],"closed-loop":[28,176],"control":[29],"when":[30],"vision-language-action":[32],"(VLA)":[33],"policy":[34,44,164],"already":[35],"uses":[36],"V-JEPA":[37,64],"features.":[38],"call":[40],"our":[41],"JEPA-conditioned":[42],"flow":[43],"family":[45],"SaiVLA-JEPA.":[46],"It":[47],"combines":[48],"vision-language":[50],"with":[52,91,96],"diffusion":[54],"transformer":[55],"action":[56,82],"expert":[57],"and":[58,75,95,137,165],"receives":[59],"features":[60],"from":[61],"frozen":[63,163],"encoder.":[65],"Two":[66],"lightweight":[67],"auxiliary":[68],"heads":[69],"future":[71,98,117],"during":[73],"training":[74,127],"are":[76],"omitted":[77],"at":[78,124],"inference,":[79],"leaving":[80],"the":[81,125,162],"sampler":[83],"unchanged.":[84],"compare":[86],"policies":[87],"without":[88,114],"JEPA":[89,93,106],"features,":[90],"current":[92],"conditioning,":[94],"additional":[97],"supervision":[99,118],"on":[100],"10,030":[101],"paired":[102,175],"LIBERO-Plus":[103],"episodes.":[104],"Current":[105],"conditioning":[107],"reaches":[108],"about":[109],"82.5%":[110],"success":[111],"versus":[112],"71.6%":[113],"JEPA,":[115],"while":[116],"provides":[119],"no":[120,158],"detectable":[121],"aggregate":[122],"gain":[123,145],"evaluated":[126],"budget.":[128],"separately":[130],"evaluate":[131],"SaiVLA-JEPA-TCA,":[132],"which":[133,139],"adapts":[134,140],"latent":[135,180],"context,":[136],"SaiVLA-JEPA-TPA,":[138],"selected":[141],"parameters.":[142],"The":[143],"context-retry":[144],"is":[146],"system-level":[148],"result":[149],"under":[150],"larger":[152],"attempt":[153],"budget;":[154],"parameter":[155],"adaptation":[156],"shows":[157],"consistent":[159],"advantage":[160],"over":[161],"target":[166],"controls.":[167],"These":[168],"results":[169],"motivate":[170],"assessing":[171],"latent-prediction":[172],"objectives":[173],"through":[174],"rollouts":[177],"loss":[181],"alone.":[182]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-31T00:00:00"}