{"id":"https://openalex.org/W7160916093","doi":"https://doi.org/10.48550/arxiv.2605.09860","title":"When to Re-Commit: Temporal Abstraction Discovery for Long-Horizon Vision-Language Reasoning","display_name":"When to Re-Commit: Temporal Abstraction Discovery for Long-Horizon Vision-Language Reasoning","publication_year":2026,"publication_date":"2026-05-11","ids":{"openalex":"https://openalex.org/W7160916093","doi":"https://doi.org/10.48550/arxiv.2605.09860"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.09860","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.09860","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.2605.09860","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135921194","display_name":"Chen Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135931326","display_name":"Zhantao Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Zhantao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135931754","display_name":"Fangyi Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Fangyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135957487","display_name":"Han Zhang","orcid":"https://orcid.org/0009-0001-2612-9166"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Han","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5001157076","display_name":"Anudeepsekhar Bolimera","orcid":"https://orcid.org/0000-0002-0068-4567"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bolimera, Anudeepsekhar","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135961405","display_name":"Marios Savvides","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Savvides, Marios","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":0.6912999749183655,"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":0.6912999749183655,"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/T10028","display_name":"Topic Modeling","score":0.06939999759197235,"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":0.04019999876618385,"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/commit","display_name":"Commit","score":0.8687999844551086},{"id":"https://openalex.org/keywords/abstraction","display_name":"Abstraction","score":0.6886000037193298},{"id":"https://openalex.org/keywords/variable","display_name":"Variable (mathematics)","score":0.4219000041484833},{"id":"https://openalex.org/keywords/compounding","display_name":"Compounding","score":0.3562000095844269},{"id":"https://openalex.org/keywords/outcome","display_name":"Outcome (game theory)","score":0.3086000084877014}],"concepts":[{"id":"https://openalex.org/C153180980","wikidata":"https://www.wikidata.org/wiki/Q19776675","display_name":"Commit","level":2,"score":0.8687999844551086},{"id":"https://openalex.org/C124304363","wikidata":"https://www.wikidata.org/wiki/Q673661","display_name":"Abstraction","level":2,"score":0.6886000037193298},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6869999766349792},{"id":"https://openalex.org/C182365436","wikidata":"https://www.wikidata.org/wiki/Q50701","display_name":"Variable (mathematics)","level":2,"score":0.4219000041484833},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4198000133037567},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.40230000019073486},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.38519999384880066},{"id":"https://openalex.org/C207673951","wikidata":"https://www.wikidata.org/wiki/Q1303150","display_name":"Compounding","level":2,"score":0.3562000095844269},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3149999976158142},{"id":"https://openalex.org/C148220186","wikidata":"https://www.wikidata.org/wiki/Q7111912","display_name":"Outcome (game theory)","level":2,"score":0.3086000084877014},{"id":"https://openalex.org/C61445026","wikidata":"https://www.wikidata.org/wiki/Q217608","display_name":"Fixed point","level":2,"score":0.2962999939918518},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2815000116825104},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.27869999408721924},{"id":"https://openalex.org/C94375191","wikidata":"https://www.wikidata.org/wiki/Q11205","display_name":"Arithmetic","level":1,"score":0.2554999887943268}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.09860","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.09860","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.2605.09860","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.09860","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Long-horizon":[0],"reasoning":[1],"requires":[2],"deciding":[3],"not":[4],"only":[5],"what":[6,86],"actions":[7,29,122],"to":[8,13,87,109],"take,":[9],"but":[10],"how":[11,91],"deeply":[12],"commit":[14],"before":[15],"the":[16,25,71,98,159,171],"next":[17],"observation.":[18],"We":[19,74,150],"formalize":[20],"this":[21,58,76],"as":[22,53,65],"\\emph{commitment":[23],"depth}:":[24],"number":[26],"of":[27,70],"primitive":[28,121],"executed":[30],"open-loop":[31],"between":[32,39],"replans.":[33],"Commitment":[34],"depth":[35,64,169,174],"induces":[36],"a":[37,54,66,78,127,153],"trade-off":[38],"replanning":[40],"cost":[41],"and":[42,89,96,134],"compounding":[43],"execution":[44],"error,":[45],"yet":[46],"most":[47],"existing":[48],"long-horizon":[49],"systems":[50],"fix":[51],"it":[52],"hand-designed":[55],"scalar.":[56],"In":[57],"work,":[59],"we":[60],"instead":[61],"treat":[62],"commitment":[63,164],"learnable,":[67],"state-conditioned":[68,163],"variable":[69],"policy":[72,81,101],"itself.":[73],"instantiate":[75],"within":[77],"model-native":[79],"vision--language":[80,144],"that":[82],"jointly":[83],"predicts":[84],"both":[85,138],"execute":[88],"for":[90],"long.":[92],"Across":[93],"Sliding":[94],"Puzzle":[95],"Sokoban,":[97],"resulting":[99],"adaptive":[100],"Pareto-dominates":[102],"every":[103,141],"non-degenerate":[104],"fixed-depth":[105],"baseline,":[106],"achieving":[107],"up":[108],"12.5":[110],"percentage":[111],"points":[112],"higher":[113],"solve":[114],"rate":[115],"while":[116,140],"using":[117,126],"approximately":[118],"25\\%":[119],"fewer":[120],"per":[123],"episode.":[124],"Despite":[125],"7B":[128],"backbone,":[129],"our":[130],"method":[131],"outperforms":[132],"GPT-5.5":[133],"Claude":[135],"Sonnet":[136],"on":[137],"tasks,":[139],"tested":[142],"open-weight":[143],"model":[145],"achieves":[146],"0\\%":[147],"zero-shot":[148],"success.":[149],"further":[151],"present":[152],"theoretical":[154],"analysis":[155],"showing":[156],"that,":[157],"under":[158],"standard":[160],"commitment-depth":[161],"surrogate,":[162],"strictly":[165],"dominates":[166],"any":[167],"fixed":[168],"whenever":[170],"locally":[172],"optimal":[173],"varies":[175],"across":[176],"states.":[177]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-13T00:00:00"}