{"id":"https://openalex.org/W7165616359","doi":"https://doi.org/10.48550/arxiv.2606.20698","title":"SafeDojo: Safe Reinforcement Learning for VLA via Interactive World Model","display_name":"SafeDojo: Safe Reinforcement Learning for VLA via Interactive World Model","publication_year":2026,"publication_date":"2026-06-15","ids":{"openalex":"https://openalex.org/W7165616359","doi":"https://doi.org/10.48550/arxiv.2606.20698"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.20698","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.20698","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.2606.20698","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139178280","display_name":"Kai Tang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Kai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112481142","display_name":"P Pengxiang Jia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jia, Peidong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139201923","display_name":"Zhong Chu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chu, Zhong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139169120","display_name":"Jixian Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Jixian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139147002","display_name":"Rui Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Rui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139134125","display_name":"Jiajun Cao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cao, Jiajun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139160381","display_name":"Fangyuan Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Fangyuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139184629","display_name":"Sixiang Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Sixiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101893660","display_name":"Yi Guo","orcid":"https://orcid.org/0009-0006-1890-6726"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Yichen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139216028","display_name":"Xiaowei Chi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chi, Xiaowei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113228249","display_name":"Chun-Kai Fan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fan, Chun-Kai","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139136995","display_name":"Kevin Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Kevin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102756789","display_name":"Jinchang Xu","orcid":"https://orcid.org/0009-0002-9838-9546"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Jinchang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139201584","display_name":"Fubing Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Fubing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139135060","display_name":"Weishi Mi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mi, Weishi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139129612","display_name":"Xiaozhu Ju","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ju, Xiaozhu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139180518","display_name":"Jian Tang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Jian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139132645","display_name":"Shanghang Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Shanghang","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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.29010000824928284,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.29010000824928284,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.20440000295639038,"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":0.09849999845027924,"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/reinforcement-learning","display_name":"Reinforcement learning","score":0.891700029373169},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.5527999997138977},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5497000217437744},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.451200008392334},{"id":"https://openalex.org/keywords/baseline","display_name":"Baseline (sea)","score":0.44209998846054077},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.43970000743865967},{"id":"https://openalex.org/keywords/learning-classifier-system","display_name":"Learning classifier system","score":0.4011000096797943}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.891700029373169},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6656000018119812},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.5527999997138977},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5497000217437744},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5153999924659729},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.451200008392334},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.44209998846054077},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.43970000743865967},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41429999470710754},{"id":"https://openalex.org/C199190896","wikidata":"https://www.wikidata.org/wiki/Q3509276","display_name":"Learning classifier system","level":3,"score":0.4011000096797943},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.4007999897003174},{"id":"https://openalex.org/C100609095","wikidata":"https://www.wikidata.org/wiki/Q1335050","display_name":"Embodied cognition","level":2,"score":0.38999998569488525},{"id":"https://openalex.org/C67203356","wikidata":"https://www.wikidata.org/wiki/Q1321905","display_name":"Reinforcement","level":2,"score":0.352400004863739},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.34700000286102295},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.3345000147819519},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.32839998602867126},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.2904999852180481},{"id":"https://openalex.org/C28006648","wikidata":"https://www.wikidata.org/wiki/Q6934509","display_name":"Multi-task learning","level":3,"score":0.28299999237060547},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.2676999866962433},{"id":"https://openalex.org/C44154836","wikidata":"https://www.wikidata.org/wiki/Q45045","display_name":"Simulation","level":1,"score":0.26429998874664307}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.20698","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.20698","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.2606.20698","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.20698","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":[{"id":"https://metadata.un.org/sdg/17","display_name":"Partnerships for the goals","score":0.4398438036441803}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Safe":[0],"control":[1],"is":[2],"a":[3,17,92,105,140,224],"prerequisite":[4],"for":[5,9,57],"real-world":[6,29],"embodied":[7,231],"intelligence,":[8],"which":[10,91],"safe":[11,22,53,63,165,220,230],"reinforcement":[12,23,54,73,221],"learning":[13,24,55,74,222],"has":[14],"emerged":[15],"as":[16,223],"promising":[18],"paradigm.":[19],"However,":[20],"existing":[21],"methods":[25],"either":[26],"require":[27],"costly":[28],"exploration":[30],"or":[31],"depend":[32],"on":[33,75,185,197],"hand-crafted":[34],"safety":[35,107,111,152],"functions.":[36],"Neither":[37],"scales":[38],"to":[39,61],"vision-language-action":[40,58],"models":[41],"deployed":[42],"in":[43],"open-world":[44],"physical":[45],"environments.":[46],"We":[47],"propose":[48],"SafeDojo,":[49],"the":[50,118,160,180,194,205],"first":[51],"model-based":[52,67,176,219],"framework":[56],"policies":[59],"designed":[60],"learn":[62],"actions":[64],"through":[65,139],"world":[66,81,84,218],"imagination.":[68],"Specifically,":[69],"SafeDojo":[70,158],"performs":[71],"online":[72],"top":[76],"of":[77,125,148],"an":[78,189],"interactive":[79],"video":[80],"model.":[82],"The":[83,131],"model":[85],"generates":[86],"action-conditioned":[87],"future":[88],"predictions,":[89],"from":[90,101,113],"tailored":[93],"ResNet":[94],"success":[95,150],"classifier":[96],"estimates":[97],"per-step":[98,110],"task":[99,126,149,163,208],"progress":[100],"imagined":[102],"frames":[103],"and":[104,128,134,151,167,175,188,209,226],"lightweight":[106],"head":[108],"predicts":[109],"costs":[112],"latent":[114],"context":[115],"together":[116],"with":[117,179],"proposed":[119],"action":[120],"chunk,":[121],"enabling":[122,145],"simultaneous":[123],"assessment":[124],"execution":[127,168],"trajectory":[129],"safety.":[130],"decoupled":[132],"task-reward":[133],"safety-cost":[135],"signals":[136],"are":[137],"balanced":[138],"Lagrangian-based":[141],"constrained":[142],"GRPO":[143],"objective,":[144],"coordinated":[146],"improvement":[147,192],"under":[153],"explicit":[154],"constraints.":[155],"On":[156],"SafeLIBERO,":[157],"achieves":[159],"best":[161,181,206],"aggregate":[162],"success,":[164,166],"efficiency":[169],"among":[170],"inference-time":[171],"safety,":[172],"model-free":[173],"RL,":[174],"RL":[177],"baselines,":[178],"average":[182,207],"safe-success":[183,210],"rate":[184],"both":[186],"levels":[187],"8.25":[190],"percentage-point":[191],"over":[193],"strongest":[195],"baseline":[196],"Level":[198],"I.":[199],"Real-world":[200],"Franka":[201],"deployment":[202],"further":[203],"shows":[204],"rates":[211],"across":[212],"five":[213],"tasks.":[214],"Our":[215],"results":[216],"position":[217],"scalable":[225],"generalizable":[227],"path":[228],"toward":[229],"intelligence.":[232]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-24T00:00:00"}