{"id":"https://openalex.org/W7166807664","doi":"https://doi.org/10.48550/arxiv.2606.31846","title":"Z-1: Efficient Reinforcement Learning for Vision-Language-Action Models","display_name":"Z-1: Efficient Reinforcement Learning for Vision-Language-Action Models","publication_year":2026,"publication_date":"2026-06-30","ids":{"openalex":"https://openalex.org/W7166807664","doi":"https://doi.org/10.48550/arxiv.2606.31846"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.31846","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.31846","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.2606.31846","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139761542","display_name":"Lang Cao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cao, Lang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122380612","display_name":"Renhong Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Renhong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139782211","display_name":"Luyi Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Luyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139761271","display_name":"Peng Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Peng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139801955","display_name":"Mofan Peng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Peng, Mofan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139722510","display_name":"Yitong Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yitong","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":0.8373000025749207,"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":0.8373000025749207,"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":0.06520000100135803,"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":0.0272000003606081,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.7936999797821045},{"id":"https://openalex.org/keywords/initialization","display_name":"Initialization","score":0.6840999722480774},{"id":"https://openalex.org/keywords/stability","display_name":"Stability (learning theory)","score":0.6137999892234802},{"id":"https://openalex.org/keywords/trajectory","display_name":"Trajectory","score":0.5447999835014343},{"id":"https://openalex.org/keywords/action","display_name":"Action (physics)","score":0.4075999855995178},{"id":"https://openalex.org/keywords/policy-learning","display_name":"Policy learning","score":0.35190001130104065},{"id":"https://openalex.org/keywords/robot","display_name":"Robot","score":0.32989999651908875}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7936999797821045},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7167999744415283},{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.6840999722480774},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.6137999892234802},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5893999934196472},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.5447999835014343},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5067999958992004},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.4075999855995178},{"id":"https://openalex.org/C2779436431","wikidata":"https://www.wikidata.org/wiki/Q30672407","display_name":"Policy learning","level":2,"score":0.35190001130104065},{"id":"https://openalex.org/C90509273","wikidata":"https://www.wikidata.org/wiki/Q11012","display_name":"Robot","level":2,"score":0.32989999651908875},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.28119999170303345},{"id":"https://openalex.org/C18555067","wikidata":"https://www.wikidata.org/wiki/Q8375051","display_name":"Joint (building)","level":2,"score":0.27320000529289246},{"id":"https://openalex.org/C77967617","wikidata":"https://www.wikidata.org/wiki/Q4677561","display_name":"Active learning (machine learning)","level":2,"score":0.27090001106262207},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.26159998774528503},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.26030001044273376},{"id":"https://openalex.org/C2781311116","wikidata":"https://www.wikidata.org/wiki/Q83306","display_name":"Group (periodic table)","level":2,"score":0.25940001010894775},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.25589999556541443}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.31846","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.31846","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.2606.31846","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.31846","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Vision-Language-Action":[0],"(VLA)":[1],"models":[2],"offer":[3],"a":[4,52,79],"promising":[5],"framework":[6,57],"for":[7,58,74],"robotic":[8],"manipulation":[9],"by":[10,25,139],"connecting":[11],"language":[12],"instructions,":[13],"visual":[14],"observations,":[15],"and":[16,76,96,112,118,142],"continuous":[17],"control.":[18],"However,":[19],"most":[20],"existing":[21],"policies":[22,160],"remain":[23],"limited":[24,37],"behavior":[26],"cloning":[27],"or":[28],"supervised":[29],"fine-tuning":[30],"(SFT)":[31],"from":[32,41],"fixed":[33],"demonstrations,":[34],"which":[35],"provides":[36],"opportunity":[38],"to":[39],"improve":[40,93,157],"the":[42,94,144],"policy's":[43],"own":[44],"failures.":[45],"In":[46],"this":[47],"paper,":[48],"we":[49],"present":[50],"Z-1,":[51],"reinforcement":[53],"learning":[54],"(RL)":[55],"post-training":[56,154],"flow-based":[59,158],"VLA":[60,159],"models.":[61,147],"Built":[62],"on":[63],"top":[64],"of":[65,98,116,132],"$\u03c0_{0.5}$,":[66],"Z-1":[67,101,126],"uses":[68],"only":[69],"publicly":[70],"released":[71],"RoboCasa":[72,90,124],"demonstrations":[73],"SFT":[75,137],"then":[77],"applies":[78],"task-wise":[80],"Group":[81],"Relative":[82],"Policy":[83],"Optimization":[84],"(GRPO)":[85],"strategy":[86],"across":[87],"$24$":[88,123],"standard":[89],"tasks.":[91],"To":[92],"efficiency":[95],"stability":[97],"online":[99],"optimization,":[100],"combines":[102],"shared-prefix":[103],"rollout":[104],"construction,":[105],"tree-structured":[106],"trajectory":[107],"branching,":[108],"completion-aware":[109],"reward":[110],"calibration,":[111],"selective":[113],"joint":[114],"training":[115],"VLM":[117],"Action":[119],"Expert.":[120],"Across":[121],"all":[122],"tasks,":[125],"achieves":[127],"an":[128],"average":[129],"success":[130],"rate":[131],"$80.6\\%$,":[133],"improving":[134],"over":[135],"its":[136],"initialization":[138],"$13.2\\%$":[140],"points":[141],"outperforms":[143],"published":[145],"sota":[146],"These":[148],"results":[149],"show":[150],"that":[151],"systematic":[152],"GRPO":[153],"can":[155],"substantially":[156],"without":[161],"additional":[162],"private":[163],"demonstrations.":[164]},"counts_by_year":[],"updated_date":"2026-07-02T06:18:51.028212","created_date":"2026-07-02T00:00:00"}