{"id":"https://openalex.org/W7161979593","doi":"https://doi.org/10.48550/arxiv.2605.20299","title":"Mechanisms of Misgeneralization in Physical Sequence Modeling","display_name":"Mechanisms of Misgeneralization in Physical Sequence Modeling","publication_year":2026,"publication_date":"2026-05-19","ids":{"openalex":"https://openalex.org/W7161979593","doi":"https://doi.org/10.48550/arxiv.2605.20299"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.20299","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.20299","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.2605.20299","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5001141874","display_name":"Kento Nishi","orcid":"https://orcid.org/0000-0002-6947-5675"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nishi, Kento","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101066425","display_name":"Raphael Tang","orcid":"https://orcid.org/0009-0007-2873-892X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Raphael","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136629776","display_name":"Karun Kumar","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kumar, Karun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5084194574","display_name":"Core Francisco Park","orcid":"https://orcid.org/0000-0002-9542-2913"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Park, Core Francisco","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136670062","display_name":"Hidenori Tanaka","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tanaka, Hidenori","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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.2867000102996826,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.2867000102996826,"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":0.06620000302791595,"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":0.061400000005960464,"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/trajectory","display_name":"Trajectory","score":0.6636999845504761},{"id":"https://openalex.org/keywords/motion","display_name":"Motion (physics)","score":0.6047000288963318},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.5407000184059143},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.5078999996185303},{"id":"https://openalex.org/keywords/physical-system","display_name":"Physical system","score":0.4864000082015991},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.4715999960899353},{"id":"https://openalex.org/keywords/kinematics","display_name":"Kinematics","score":0.44859999418258667},{"id":"https://openalex.org/keywords/robotics","display_name":"Robotics","score":0.41659998893737793},{"id":"https://openalex.org/keywords/aggregate","display_name":"Aggregate (composite)","score":0.41600000858306885}],"concepts":[{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.6636999845504761},{"id":"https://openalex.org/C104114177","wikidata":"https://www.wikidata.org/wiki/Q79782","display_name":"Motion (physics)","level":2,"score":0.6047000288963318},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.604200005531311},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5885000228881836},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.5407000184059143},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.5078999996185303},{"id":"https://openalex.org/C116672817","wikidata":"https://www.wikidata.org/wiki/Q1454986","display_name":"Physical system","level":2,"score":0.4864000082015991},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.4715999960899353},{"id":"https://openalex.org/C39920418","wikidata":"https://www.wikidata.org/wiki/Q11476","display_name":"Kinematics","level":2,"score":0.44859999418258667},{"id":"https://openalex.org/C34413123","wikidata":"https://www.wikidata.org/wiki/Q170978","display_name":"Robotics","level":3,"score":0.41659998893737793},{"id":"https://openalex.org/C4679612","wikidata":"https://www.wikidata.org/wiki/Q866298","display_name":"Aggregate (composite)","level":2,"score":0.41600000858306885},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.3880999982357025},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.38359999656677246},{"id":"https://openalex.org/C527412718","wikidata":"https://www.wikidata.org/wiki/Q855395","display_name":"Interpretation (philosophy)","level":2,"score":0.34940001368522644},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.3483000099658966},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.3296000063419342},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3183000087738037},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.30329999327659607},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.30320000648498535},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.29989999532699585},{"id":"https://openalex.org/C2776505523","wikidata":"https://www.wikidata.org/wiki/Q4785468","display_name":"Plan (archaeology)","level":2,"score":0.29829999804496765},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.29280000925064087},{"id":"https://openalex.org/C44154836","wikidata":"https://www.wikidata.org/wiki/Q45045","display_name":"Simulation","level":1,"score":0.28929999470710754},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.2881999909877777},{"id":"https://openalex.org/C40506919","wikidata":"https://www.wikidata.org/wiki/Q7452469","display_name":"Sequence learning","level":2,"score":0.28679999709129333},{"id":"https://openalex.org/C48007421","wikidata":"https://www.wikidata.org/wiki/Q676252","display_name":"Motion capture","level":3,"score":0.27630001306533813},{"id":"https://openalex.org/C90509273","wikidata":"https://www.wikidata.org/wiki/Q11012","display_name":"Robot","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C2780428219","wikidata":"https://www.wikidata.org/wiki/Q16952335","display_name":"Cover (algebra)","level":2,"score":0.2590000033378601},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2563000023365021},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2547000050544739}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.20299","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.20299","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.2605.20299","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.20299","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":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Generative":[0],"sequence":[1],"models":[2],"are":[3,189],"often":[4],"trained":[5],"to":[6,14,21,30,68,140,158,196],"plan":[7],"motion":[8,178],"in":[9,167],"physical":[10,39,100,108,124,138,161],"domains,":[11],"from":[12],"robotics":[13],"mechanical":[15,45],"simulations.":[16],"When":[17],"constructing":[18],"a":[19,24,38,49,52,63,150,198],"dataset":[20],"train":[22],"such":[23],"model,":[25],"engineers":[26],"may":[27],"curate":[28],"demonstrations":[29,58],"specify":[31],"how":[32],"trajectories":[33],"should":[34],"be":[35],"distributed":[36],"over":[37,98],"quantity":[40,101],"like":[41],"travel":[42,60],"distance":[43],"or":[44,164],"energy.":[46],"For":[47],"example,":[48],"roboticist":[50],"building":[51],"maze":[53,174],"navigation":[54,175],"agent":[55],"might":[56],"choose":[57],"whose":[59],"distances":[61],"cover":[62],"fixed":[64],"range":[65],"uniformly,":[66],"hoping":[67],"constrain":[69],"the":[70,95,99,132,137,142],"agent's":[71],"expected":[72],"power":[73],"usage.":[74],"We":[75,104,145],"find":[76],"that":[77,123],"standard":[78],"deep":[79],"learning":[80],"can":[81,88],"violate":[82],"this":[83,106],"intent:":[84],"each":[85],"generated":[86],"trajectory":[87],"seem":[89],"plausible":[90],"on":[91],"its":[92,115],"own,":[93],"but":[94],"aggregate":[96],"distribution":[97],"is":[102],"wrong.":[103],"call":[105],"failure":[107],"misgeneralization,":[109],"and":[110,154,171,176,192],"develop":[111],"an":[112],"account":[113],"of":[114,131],"mechanism.":[116],"Using":[117],"controlled":[118],"synthetic":[119,170],"tasks,":[120],"we":[121,155,193],"show":[122],"misgeneralization":[125],"arises":[126],"when":[127],"local":[128],"errors":[129,148],"typical":[130],"model":[133],"class":[134],"propagate":[135],"through":[136],"measurement":[139],"shift":[141],"recovered":[143],"distribution.":[144],"estimate":[146],"these":[147],"with":[149],"data":[151],"deviation":[152],"kernel,":[153],"use":[156,194],"it":[157,195],"predict":[159],"which":[160,186],"quantities":[162],"gain":[163],"lose":[165],"mass":[166],"both":[168],"our":[169,181],"more":[172],"applied":[173],"double-pendulum":[177],"tasks.":[179],"Finally,":[180],"mechanistic":[182],"interpretation":[183],"helps":[184],"identify":[185],"mitigation":[187],"strategies":[188],"structurally":[190],"promising,":[191],"propose":[197],"kernel-informed":[199],"intervention.":[200]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-22T00:00:00"}