{"id":"https://openalex.org/W7163165227","doi":"https://doi.org/10.48550/arxiv.2606.01810","title":"Token Predictors Are Not Planners: Building Physically Grounded Causal Reasoners","display_name":"Token Predictors Are Not Planners: Building Physically Grounded Causal Reasoners","publication_year":2026,"publication_date":"2026-06-01","ids":{"openalex":"https://openalex.org/W7163165227","doi":"https://doi.org/10.48550/arxiv.2606.01810"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.01810","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01810","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.2606.01810","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137695935","display_name":"Zheng Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Zheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137664122","display_name":"Mingqi Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Mingqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137627339","display_name":"Qinlei Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Qinlei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101215538","display_name":"Wanqi Zhong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhong, Wanqi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137632812","display_name":"Hanwen Cui","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cui, Hanwen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101453928","display_name":"Heng Cao","orcid":"https://orcid.org/0000-0001-6160-3997"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cao, Heng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137636376","display_name":"Zirui Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Zirui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137618639","display_name":"Yifan Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Yifan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137615311","display_name":"Chong Luo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Chong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137687446","display_name":"Bei Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Bei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137619699","display_name":"Yiming Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yiming","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":0.8669999837875366,"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":0.8669999837875366,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.02759999968111515,"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":0.013100000098347664,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.6349999904632568},{"id":"https://openalex.org/keywords/causal-model","display_name":"Causal model","score":0.5654000043869019},{"id":"https://openalex.org/keywords/embodied-cognition","display_name":"Embodied cognition","score":0.5601999759674072},{"id":"https://openalex.org/keywords/causal-reasoning","display_name":"Causal reasoning","score":0.5167999863624573},{"id":"https://openalex.org/keywords/causality","display_name":"Causality (physics)","score":0.508400022983551},{"id":"https://openalex.org/keywords/causal-inference","display_name":"Causal inference","score":0.47940000891685486},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.4715999960899353},{"id":"https://openalex.org/keywords/grounded-theory","display_name":"Grounded theory","score":0.46560001373291016},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.4381999969482422}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6395999789237976},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.6349999904632568},{"id":"https://openalex.org/C11671645","wikidata":"https://www.wikidata.org/wiki/Q5054567","display_name":"Causal model","level":2,"score":0.5654000043869019},{"id":"https://openalex.org/C100609095","wikidata":"https://www.wikidata.org/wiki/Q1335050","display_name":"Embodied cognition","level":2,"score":0.5601999759674072},{"id":"https://openalex.org/C115086926","wikidata":"https://www.wikidata.org/wiki/Q17004651","display_name":"Causal reasoning","level":3,"score":0.5167999863624573},{"id":"https://openalex.org/C64357122","wikidata":"https://www.wikidata.org/wiki/Q1149766","display_name":"Causality (physics)","level":2,"score":0.508400022983551},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5001000165939331},{"id":"https://openalex.org/C158600405","wikidata":"https://www.wikidata.org/wiki/Q5054566","display_name":"Causal inference","level":2,"score":0.47940000891685486},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.4715999960899353},{"id":"https://openalex.org/C156325361","wikidata":"https://www.wikidata.org/wiki/Q1152864","display_name":"Grounded theory","level":3,"score":0.46560001373291016},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.4381999969482422},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.41780000925064087},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.37389999628067017},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.35989999771118164},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.32269999384880066},{"id":"https://openalex.org/C188147891","wikidata":"https://www.wikidata.org/wiki/Q147638","display_name":"Cognitive science","level":1,"score":0.31459999084472656},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.3001999855041504},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.2957000136375427},{"id":"https://openalex.org/C23224414","wikidata":"https://www.wikidata.org/wiki/Q176769","display_name":"Hidden Markov model","level":2,"score":0.29319998621940613},{"id":"https://openalex.org/C97256817","wikidata":"https://www.wikidata.org/wiki/Q1462316","display_name":"Spurious relationship","level":2,"score":0.29179999232292175},{"id":"https://openalex.org/C65414064","wikidata":"https://www.wikidata.org/wiki/Q484105","display_name":"Autonomy","level":2,"score":0.2903999984264374},{"id":"https://openalex.org/C2776321320","wikidata":"https://www.wikidata.org/wiki/Q857525","display_name":"Annotation","level":2,"score":0.2872999906539917},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.27970001101493835},{"id":"https://openalex.org/C11693617","wikidata":"https://www.wikidata.org/wiki/Q181839","display_name":"Pragmatics","level":2,"score":0.2773999869823456},{"id":"https://openalex.org/C176777502","wikidata":"https://www.wikidata.org/wiki/Q4774623","display_name":"Anticipation (artificial intelligence)","level":2,"score":0.2768000066280365},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.27489998936653137},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.27079999446868896},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.2667999863624573},{"id":"https://openalex.org/C60048249","wikidata":"https://www.wikidata.org/wiki/Q37437","display_name":"Syntax","level":2,"score":0.2533000111579895},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.2524999976158142},{"id":"https://openalex.org/C71889745","wikidata":"https://www.wikidata.org/wiki/Q1783264","display_name":"Counterfactual conditional","level":3,"score":0.25220000743865967},{"id":"https://openalex.org/C108650721","wikidata":"https://www.wikidata.org/wiki/Q1783253","display_name":"Counterfactual thinking","level":2,"score":0.25060001015663147}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.01810","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01810","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.2606.01810","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01810","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":[{"score":0.4520410895347595,"display_name":"Sustainable cities and communities","id":"https://metadata.un.org/sdg/11"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Current":[0],"benchmarks":[1],"for":[2,135],"embodied":[3,71],"vision-language":[4],"planning":[5,31,72],"often":[6],"favor":[7],"linguistic":[8],"next-token":[9],"prediction":[10,49],"over":[11,94],"physically":[12,51,187],"grounded":[13,47,52,188],"next-state":[14,138],"reasoning.":[15,54],"This":[16],"rewards":[17],"models":[18,102],"that":[19,38,100],"mimic":[20],"statistical":[21],"language":[22],"priors":[23],"rather":[24],"than":[25],"track":[26],"causal":[27,53,75,156,189],"dependencies,":[28],"reducing":[29],"physical":[30,40,108,133],"to":[32,69,105,131,159,170],"shallow":[33],"sequence":[34],"modeling.":[35],"We":[36,77],"argue":[37],"reliable":[39],"autonomy":[41],"requires":[42],"a":[43,61,81,90,151,164,176],"shift":[44],"from":[45,168,182],"linguistically":[46],"token":[48,184],"toward":[50,179],"To":[55],"this":[56],"end,":[57],"we":[58],"introduce":[59],"Causal-Plan-Bench,":[60],"high-fidelity":[62],"diagnostic":[63],"suite":[64],"curated":[65],"through":[66],"multi-stage":[67],"verification":[68],"evaluate":[70],"across":[73],"four":[74],"dimensions.":[76],"also":[78],"construct":[79],"Causal-Plan-1M,":[80],"million-scale":[82],"corpus":[83],"of":[84],"explicit":[85],"reasoning":[86],"traces":[87],"produced":[88],"by":[89],"four-stage":[91],"annotation":[92],"pipeline":[93],"egocentric":[95],"videos.":[96],"Extensive":[97],"evaluation":[98],"shows":[99],"leading":[101],"still":[103],"struggle":[104],"demonstrate":[106],"genuine":[107],"agency,":[109],"with":[110],"Gemini":[111],"3":[112],"Pro":[113],"reaching":[114],"only":[115],"38.18":[116],"on":[117,129],"our":[118,122,173],"benchmark.":[119],"In":[120],"contrast,":[121],"training":[123,157],"recipe":[124],"enables":[125],"Causal":[126,152],"Planner,":[127],"built":[128],"Qwen3-VL-8B,":[130],"internalize":[132],"logic":[134],"more":[136],"accurate":[137],"estimation.":[139],"The":[140],"model":[141],"achieves":[142],"strong":[143],"in-domain":[144],"performance":[145],"and":[146,149],"cross-benchmark":[147],"generalization,":[148],"reveals":[150],"Scaling":[153],"Law:":[154],"scaling":[155],"data":[158],"one":[160],"million":[161],"instances":[162],"yields":[163],"36.3%":[165],"relative":[166],"gain,":[167],"33.22":[169],"45.28.":[171],"Overall,":[172],"work":[174],"provides":[175],"concrete":[177],"step":[178],"turning":[180],"agents":[181],"superficial":[183],"predictors":[185],"into":[186],"reasoners.":[190]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-03T00:00:00"}