{"id":"https://openalex.org/W7204202063","doi":"https://doi.org/10.48550/arxiv.2608.22197","title":"On the Capability Separation Between World-Model Policy Learning and Imitated World-Action Models","display_name":"On the Capability Separation Between World-Model Policy Learning and Imitated World-Action Models","publication_year":2026,"publication_date":"2026-08-23","ids":{"openalex":"https://openalex.org/W7204202063","doi":"https://doi.org/10.48550/arxiv.2608.22197"},"language":null,"primary_location":{"id":"pmh:oai:arXiv.org:2608.22197","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.22197","pdf_url":"https://arxiv.org/pdf/2608.22197","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"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.22197","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5148084561","display_name":"Yang Yu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Yu, Yang","raw_affiliation_strings":["Nanjing University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University","institution_ids":["https://openalex.org/I881766915"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5148084561"],"corresponding_institution_ids":["https://openalex.org/I881766915"],"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":0.9994999766349792,"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":0.9994999766349792,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":9.999999747378752e-05,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.0,"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/world-action-models","display_name":"world action models","score":0.7609999775886536},{"id":"https://openalex.org/keywords/world-models","display_name":"world models","score":0.7490000128746033},{"id":"https://openalex.org/keywords/behavioral-cloning","display_name":"behavior cloning","score":0.6650000214576721},{"id":"https://openalex.org/keywords/imitation-learning","display_name":"imitation learning","score":0.9010000228881836},{"id":"https://openalex.org/keywords/future-prediction","display_name":"future prediction","score":0.5709999799728394},{"id":"https://openalex.org/keywords/action-effects","display_name":"action effects","score":0.335999995470047},{"id":"https://openalex.org/keywords/worst-case-regret","display_name":"worst-case regret","score":0.7229999899864197},{"id":"https://openalex.org/keywords/policy-optimization","display_name":"policy optimization","score":0.48899999260902405},{"id":"https://openalex.org/keywords/realizability","display_name":"realizability","score":0.32100000977516174}],"concepts":[{"id":"https://openalex.org/C23131810","wikidata":"https://www.wikidata.org/wiki/Q818574","display_name":"Observational study","level":2,"score":0.7358999848365784},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5964000225067139},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5756999850273132},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5464000105857849},{"id":"https://openalex.org/C102483320","wikidata":"https://www.wikidata.org/wiki/Q1143725","display_name":"Observational learning","level":3,"score":0.5426999926567078},{"id":"https://openalex.org/C126388530","wikidata":"https://www.wikidata.org/wiki/Q1131737","display_name":"Imitation","level":2,"score":0.5041999816894531},{"id":"https://openalex.org/C2908647359","wikidata":"https://www.wikidata.org/wiki/Q2625603","display_name":"Population","level":2,"score":0.47850000858306885},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.4284999966621399},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.4174000024795532},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.3961000144481659},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.37599998712539673},{"id":"https://openalex.org/C106306483","wikidata":"https://www.wikidata.org/wiki/Q183984","display_name":"Futures contract","level":2,"score":0.3619999885559082},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.33320000767707825},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.3109000027179718},{"id":"https://openalex.org/C187834632","wikidata":"https://www.wikidata.org/wiki/Q188804","display_name":"Factorization","level":2,"score":0.30630001425743103},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.3059999942779541},{"id":"https://openalex.org/C148220186","wikidata":"https://www.wikidata.org/wiki/Q7111912","display_name":"Outcome (game theory)","level":2,"score":0.3034000098705292},{"id":"https://openalex.org/C12267149","wikidata":"https://www.wikidata.org/wiki/Q282453","display_name":"Support vector machine","level":2,"score":0.289000004529953},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.2865999937057495},{"id":"https://openalex.org/C2780665704","wikidata":"https://www.wikidata.org/wiki/Q959298","display_name":"Intervention (counseling)","level":2,"score":0.2854999899864197},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.27480000257492065},{"id":"https://openalex.org/C121050878","wikidata":"https://www.wikidata.org/wiki/Q5135020","display_name":"Cloning (programming)","level":2,"score":0.2694999873638153},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.2685999870300293},{"id":"https://openalex.org/C774472","wikidata":"https://www.wikidata.org/wiki/Q6760393","display_name":"Margin (machine learning)","level":2,"score":0.2671999931335449},{"id":"https://openalex.org/C2776639384","wikidata":"https://www.wikidata.org/wiki/Q840396","display_name":"Ideal (ethics)","level":2,"score":0.26570001244544983},{"id":"https://openalex.org/C2776502983","wikidata":"https://www.wikidata.org/wiki/Q690182","display_name":"Contrast (vision)","level":2,"score":0.2603999972343445},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.2551000118255615}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.22197","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.22197","pdf_url":"https://arxiv.org/pdf/2608.22197","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.2608.22197","doi":"https://doi.org/10.48550/arxiv.2608.22197","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.22197","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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.22197","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.22197","pdf_url":"https://arxiv.org/pdf/2608.22197","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":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204202063.pdf","grobid_xml":"https://content.openalex.org/works/W7204202063.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"World-action":[0],"models":[1,154],"predict":[2],"a":[3,46,55,74,142,167],"future":[4,111,153],"outcome":[5],"and":[6,19,54,95,101,138,176,222],"then":[7],"infer":[8],"an":[9,50,59,172,191],"associated":[10,218],"action.":[11],"Although":[12],"this":[13],"factorization":[14,116],"can":[15,70,170],"improve":[16],"representation":[17],"learning":[18,115,130],"data":[20],"efficiency,":[21],"it":[22,26],"is":[23,213],"unclear":[24],"whether":[25],"provides":[27],"stronger":[28],"control":[29,143],"capability":[30],"than":[31],"direct":[32,47,75,98],"behavior":[33,99,108,221],"cloning":[34,100],"when":[35],"both":[36,104],"are":[37],"trained":[38],"from":[39],"the":[40,64,79,85,106,114,119,147,160],"same":[41,80],"observational":[42,107,179,197],"demonstrations.":[43],"We":[44,145],"compare":[45],"behavior-cloning":[48],"policy,":[49,53],"imitation-trained":[51],"world-action":[52,68,102,168],"policy":[56,69,77,122,229],"optimized":[57],"with":[58,78,219],"action-conditioned":[60],"world":[61],"model.":[62],"At":[63,84],"controller-class":[65],"level,":[66,87],"every":[67,196],"be":[71],"flattened":[72],"into":[73],"stochastic":[76],"closed-loop":[81],"trajectory":[82],"distribution.":[83],"population":[86],"under":[88,135,165],"realizability,":[89],"exact":[90],"optimization,":[91],"common":[92],"deployment":[93],"information,":[94],"distribution-preserving":[96],"deployment,":[97],"imitation":[103,126],"recover":[105,171],"policy.":[109],"Thus,":[110],"prediction":[112,150],"changes":[113],"but":[117],"not":[118,157,182],"unrestricted":[120],"external":[121],"class":[123],"or":[124],"ideal":[125],"target.":[127],"Action-conditioned":[128],"world-model":[129],"differs":[131],"by":[132],"predicting":[133,216,223],"outcomes":[134],"specified":[136,226],"actions":[137,227],"comparing":[139],"them":[140],"through":[141],"objective.":[144],"characterize":[146],"irreducible":[148],"action-specific":[149],"error":[151],"of":[152,225],"that":[155,178],"do":[156,181],"condition":[158],"on":[159],"candidate":[161],"action,":[162],"identify":[163,183],"conditions":[164],"which":[166,195],"joint":[169],"interventional":[173],"forward":[174],"model,":[175],"show":[177],"demonstrations":[180],"action":[184],"effects":[185],"in":[186,194],"general.":[187],"Finally,":[188],"we":[189],"construct":[190],"environment":[192],"family":[193],"learner":[198],"has":[199],"positive":[200],"worst-case":[201],"regret,":[202],"whereas":[203],"one":[204],"informative":[205],"intervention":[206],"permits":[207],"zero":[208],"regret.":[209],"The":[210],"key":[211],"distinction":[212],"therefore":[214],"between":[215],"futures":[217],"observed":[220],"consequences":[224],"for":[228],"optimization.":[230]},"counts_by_year":[],"updated_date":"2026-09-30T09:27:27.532063","created_date":"2026-08-26T00:00:00"}