{"id":"https://openalex.org/W7165388901","doi":"https://doi.org/10.48550/arxiv.2606.19919","title":"ADaPT: Token-Level Decoupling for Efficient Large Reasoning Models","display_name":"ADaPT: Token-Level Decoupling for Efficient Large Reasoning Models","publication_year":2026,"publication_date":"2026-06-18","ids":{"openalex":"https://openalex.org/W7165388901","doi":"https://doi.org/10.48550/arxiv.2606.19919"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.19919","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19919","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":null,"license_id":null,"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.2606.19919","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5043395445","display_name":"Tingyun Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Tingyun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101249218","display_name":"Zishang Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiang, Zishang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139006843","display_name":"Jinyi Han","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han, Jinyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139000379","display_name":"Xinyi Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Xinyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138965410","display_name":"Sihang Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiang, Sihang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100722203","display_name":"Han Xia","orcid":"https://orcid.org/0000-0002-5179-0255"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xia, Han","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139005885","display_name":"Zhaoqian Dai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dai, Zhaoqian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138979531","display_name":"Shuguang Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Shuguang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138995053","display_name":"Fei Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Fei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138988267","display_name":"Jiaqing Liang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liang, Jiaqing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133363812","display_name":"Y H Xiao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiao, Yanghua","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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.15839999914169312,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.15839999914169312,"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/T11303","display_name":"Bayesian Modeling and Causal Inference","score":0.15610000491142273,"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":0.13940000534057617,"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/correctness","display_name":"Correctness","score":0.7560999989509583},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6054999828338623},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.5408999919891357},{"id":"https://openalex.org/keywords/reasoning-system","display_name":"Reasoning system","score":0.44020000100135803},{"id":"https://openalex.org/keywords/qualitative-reasoning","display_name":"Qualitative reasoning","score":0.4293000102043152},{"id":"https://openalex.org/keywords/decoupling","display_name":"Decoupling (probability)","score":0.415800005197525},{"id":"https://openalex.org/keywords/control","display_name":"Control (management)","score":0.41519999504089355}],"concepts":[{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.7560999989509583},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7229999899864197},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6054999828338623},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.5408999919891357},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.45010000467300415},{"id":"https://openalex.org/C89288958","wikidata":"https://www.wikidata.org/wiki/Q7301504","display_name":"Reasoning system","level":2,"score":0.44020000100135803},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.4293000102043152},{"id":"https://openalex.org/C205606062","wikidata":"https://www.wikidata.org/wiki/Q5249645","display_name":"Decoupling (probability)","level":2,"score":0.415800005197525},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.41519999504089355},{"id":"https://openalex.org/C37335422","wikidata":"https://www.wikidata.org/wiki/Q6888134","display_name":"Model-based reasoning","level":3,"score":0.3950999975204468},{"id":"https://openalex.org/C2986314615","wikidata":"https://www.wikidata.org/wiki/Q36829","display_name":"Pareto optimal","level":3,"score":0.3781999945640564},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.3714999854564667},{"id":"https://openalex.org/C159032336","wikidata":"https://www.wikidata.org/wiki/Q2488768","display_name":"Non-monotonic logic","level":2,"score":0.34779998660087585},{"id":"https://openalex.org/C86827895","wikidata":"https://www.wikidata.org/wiki/Q7098582","display_name":"Opportunistic reasoning","level":4,"score":0.3301999866962433},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3237999975681305},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.3066999912261963}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.19919","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19919","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.19919","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.19919","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":null,"license_id":null,"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":{"Large":[0],"reasoning":[1,14,28,33,55,103,156],"models":[2],"rely":[3],"on":[4],"long":[5,52,102],"chain-of-thought":[6],"to":[7,24,85,95,98],"achieve":[8],"strong":[9,155],"performance,":[10],"but":[11,53],"applying":[12,91],"such":[13],"uniformly":[15],"incurs":[16],"high":[17],"computational":[18],"cost.":[19],"Existing":[20],"efficiency-oriented":[21],"methods":[22],"attempt":[23],"shorten":[25],"or":[26],"mix":[27],"strategies,":[29],"yet":[30],"often":[31],"degrade":[32],"capability.":[34],"We":[35],"identify":[36],"the":[37,117,125,129,140],"root":[38],"cause":[39],"as":[40],"sequence-level":[41],"coupling":[42],"between":[43],"efficiency":[44,74,106],"incentives":[45],"and":[46,75,88,113],"correctness":[47,76],"optimization,":[48],"which":[49],"implicitly":[50],"penalizes":[51],"correct":[54,101],"trajectories.":[56],"To":[57],"address":[58],"this":[59,96],"issue,":[60],"we":[61],"propose":[62],"Adaptive":[63],"Dual-Process":[64],"Thinking":[65],"(ADaPT),":[66],"a":[67,82,132],"token-level":[68],"dual-process":[69],"framework":[70],"that":[71,147],"explicitly":[72],"decouples":[73],"signals":[77],"during":[78],"training.":[79],"ADaPT":[80,110,148],"introduces":[81],"mode-selection":[83,130],"token":[84,97],"control":[86,115],"fast":[87],"slow":[89],"reasoning,":[90],"efficiency-related":[92],"rewards":[93],"exclusively":[94],"avoid":[99],"penalizing":[100],"while":[104,153],"encouraging":[105],"when":[107],"appropriate.":[108],"Moreover,":[109],"enables":[111],"precise":[112],"continuous":[114],"over":[116],"efficiency-performance":[118,141],"trade-off":[119],"at":[120],"inference":[121,151],"time:":[122],"by":[123],"adjusting":[124],"generation":[126],"probability":[127],"of":[128],"token,":[131],"single":[133],"trained":[134],"model":[135],"can":[136],"smoothly":[137],"move":[138],"along":[139],"Pareto":[142],"frontier.":[143],"Extensive":[144],"experiments":[145],"demonstrate":[146],"significantly":[149],"reduces":[150],"cost":[152],"maintaining":[154],"performance":[157],"across":[158],"multiple":[159],"benchmarks.":[160]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-20T00:00:00"}