{"id":"https://openalex.org/W7127888457","doi":"https://doi.org/10.48550/arxiv.2602.05842","title":"Reinforcement World Model Learning for LLM-based Agents","display_name":"Reinforcement World Model Learning for LLM-based Agents","publication_year":2026,"publication_date":"2026-02-05","ids":{"openalex":"https://openalex.org/W7127888457","doi":"https://doi.org/10.48550/arxiv.2602.05842"},"language":null,"primary_location":{"id":"pmh:doi:10.48550/arxiv.2602.05842","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":"Article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":null,"any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5125179825","display_name":"Xiao Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Xiao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066404470","display_name":"Baolin Peng","orcid":"https://orcid.org/0009-0004-3255-8735"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Peng, Baolin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125109384","display_name":"Ruize Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Ruize","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125187616","display_name":"Yelong Shen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shen, Yelong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125128722","display_name":"Pengcheng He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Pengcheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024224291","display_name":"Suman Nath","orcid":"https://orcid.org/0000-0001-7813-9756"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nath, Suman","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125164093","display_name":"Nikhil Singh","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Singh, Nikhil","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089871459","display_name":"Jiangfeng Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Jiangfeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5125111181","display_name":"Zhou Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Zhou","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"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/T10028","display_name":"Topic Modeling","score":0.3249000012874603,"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/T10028","display_name":"Topic Modeling","score":0.3249000012874603,"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":0.15440000593662262,"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/T13702","display_name":"Machine Learning in Healthcare","score":0.06639999896287918,"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":0.7545999884605408},{"id":"https://openalex.org/keywords/fidelity","display_name":"Fidelity","score":0.5794000029563904},{"id":"https://openalex.org/keywords/consistency","display_name":"Consistency (knowledge bases)","score":0.5526000261306763},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.4575999975204468},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.44690001010894775},{"id":"https://openalex.org/keywords/action","display_name":"Action (physics)","score":0.4203000068664551},{"id":"https://openalex.org/keywords/equivalence","display_name":"Equivalence (formal languages)","score":0.36809998750686646},{"id":"https://openalex.org/keywords/base","display_name":"Base (topology)","score":0.3098999857902527}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.7545999884605408},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6830999851226807},{"id":"https://openalex.org/C2776459999","wikidata":"https://www.wikidata.org/wiki/Q2119376","display_name":"Fidelity","level":2,"score":0.5794000029563904},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5740000009536743},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.5526000261306763},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.4575999975204468},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.44690001010894775},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.4203000068664551},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.36809998750686646},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.34709998965263367},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.3098999857902527},{"id":"https://openalex.org/C28427503","wikidata":"https://www.wikidata.org/wiki/Q13580300","display_name":"Internal model","level":3,"score":0.3050000071525574},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.30399999022483826},{"id":"https://openalex.org/C166109690","wikidata":"https://www.wikidata.org/wiki/Q4677422","display_name":"Action selection","level":3,"score":0.2865000069141388},{"id":"https://openalex.org/C97508593","wikidata":"https://www.wikidata.org/wiki/Q1359069","display_name":"Contingency","level":2,"score":0.28619998693466187},{"id":"https://openalex.org/C196340769","wikidata":"https://www.wikidata.org/wiki/Q7698910","display_name":"Temporal difference learning","level":3,"score":0.28029999136924744},{"id":"https://openalex.org/C185874996","wikidata":"https://www.wikidata.org/wiki/Q269699","display_name":"Interdependence","level":2,"score":0.2732999920845032},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.26420000195503235},{"id":"https://openalex.org/C45493050","wikidata":"https://www.wikidata.org/wiki/Q7884934","display_name":"Unified Model","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.26170000433921814},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.25540000200271606}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:doi:10.48550/arxiv.2602.05842","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":"Article"},{"id":"doi:10.48550/arxiv.2602.05842","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2602.05842","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":"article"}],"best_oa_location":{"id":"pmh:doi:10.48550/arxiv.2602.05842","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"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":"Article"},"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":{"Large":[0],"language":[1],"models":[2,50],"(LLMs)":[3],"have":[4],"achieved":[5],"strong":[6],"performance":[7,180],"in":[8,12,33,89],"language-centric":[9],"tasks.":[10],"However,":[11],"agentic":[13],"settings,":[14],"LLMs":[15],"often":[16],"struggle":[17],"to":[18,24,112,128],"anticipate":[19],"action":[20],"consequences":[21],"and":[22,85,109,123,139,142,168,173],"adapt":[23],"environment":[25,87],"dynamics,":[26],"highlighting":[27],"the":[28,69,77,147,179],"need":[29],"for":[30,51],"world-modeling":[31],"capabilities":[32],"LLM-based":[34,52],"agents.":[35],"We":[36,133],"propose":[37],"Reinforcement":[38],"World":[39],"Model":[40],"Learning":[41],"(RWML),":[42],"a":[43,90,118],"self-supervised":[44],"method":[45,62,116,136,160],"that":[46],"learns":[47],"action-conditioned":[48],"world":[49,83],"agents":[53],"on":[54,137,171],"textual":[55],"states":[56,66,74],"using":[57],"sim-to-real":[58],"gap":[59],"rewards.":[60],"Our":[61],"aligns":[63],"simulated":[64],"next":[65,73],"produced":[67],"by":[68,166],"model":[70,113],"with":[71,156],"realized":[72],"observed":[75],"from":[76],"environment,":[78],"encouraging":[79],"consistency":[80],"between":[81],"internal":[82],"simulations":[84],"actual":[86],"dynamics":[88],"pre-trained":[91],"embedding":[92],"space.":[93],"Unlike":[94],"next-state":[95],"token":[96],"prediction,":[97],"which":[98],"prioritizes":[99],"token-level":[100],"fidelity":[101],"(i.e.,":[102],"reproducing":[103],"exact":[104],"wording)":[105],"over":[106,146],"semantic":[107],"equivalence":[108],"can":[110],"lead":[111],"collapse,":[114],"our":[115,135,159],"provides":[117],"more":[119],"robust":[120],"training":[121],"signal":[122],"is":[124],"empirically":[125],"less":[126],"susceptible":[127],"reward":[129,164],"hacking":[130],"than":[131],"LLM-as-a-judge.":[132],"evaluate":[134],"ALFWorld":[138,172],"$\u03c4^2$":[140,174],"Bench":[141,175],"observe":[143],"significant":[144],"gains":[145],"base":[148],"model,":[149],"despite":[150],"being":[151],"entirely":[152],"self-supervised.":[153],"When":[154],"combined":[155],"task-success":[157,163],"rewards,":[158],"outperforms":[161],"direct":[162],"RL":[165],"6.9":[167],"5.7":[169],"points":[170],"respectively,":[176],"while":[177],"matching":[178],"of":[181],"expert-data":[182],"training.":[183]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-02-07T00:00:00"}