{"id":"https://openalex.org/W7152611113","doi":"https://doi.org/10.48550/arxiv.2604.06787","title":"When Is Thinking Enough? Early Exit via Sufficiency Assessment for Efficient Reasoning","display_name":"When Is Thinking Enough? Early Exit via Sufficiency Assessment for Efficient Reasoning","publication_year":2026,"publication_date":"2026-04-08","ids":{"openalex":"https://openalex.org/W7152611113","doi":"https://doi.org/10.48550/arxiv.2604.06787"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.06787","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.06787","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2604.06787","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5133301861","display_name":"Yang Xiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiang, Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133251304","display_name":"Yixin Ji","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji, Yixin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037627177","display_name":"Ruotao Xu","orcid":"https://orcid.org/0000-0002-5277-9859"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Ruotao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133309207","display_name":"Dan Qiao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qiao, Dan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133272829","display_name":"Zheming Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Zheming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5133285676","display_name":"Juntao Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Juntao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133307825","display_name":"Min Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Min","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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.08560000360012054,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.08560000360012054,"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/T13219","display_name":"Mind wandering and attention","score":0.07429999858140945,"subfield":{"id":"https://openalex.org/subfields/2805","display_name":"Cognitive Neuroscience"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.07289999723434448,"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/redundancy","display_name":"Redundancy (engineering)","score":0.5537999868392944},{"id":"https://openalex.org/keywords/overconfidence-effect","display_name":"Overconfidence effect","score":0.5357999801635742},{"id":"https://openalex.org/keywords/reflection","display_name":"Reflection (computer programming)","score":0.5285999774932861},{"id":"https://openalex.org/keywords/point","display_name":"Point (geometry)","score":0.4120999872684479},{"id":"https://openalex.org/keywords/analytic-reasoning","display_name":"Analytic reasoning","score":0.39879998564720154},{"id":"https://openalex.org/keywords/reasoning-system","display_name":"Reasoning system","score":0.35359999537467957},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.3361999988555908}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6288999915122986},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.5537999868392944},{"id":"https://openalex.org/C51110983","wikidata":"https://www.wikidata.org/wiki/Q16503490","display_name":"Overconfidence effect","level":2,"score":0.5357999801635742},{"id":"https://openalex.org/C65682993","wikidata":"https://www.wikidata.org/wiki/Q1056451","display_name":"Reflection (computer programming)","level":2,"score":0.5285999774932861},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5023000240325928},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.4120999872684479},{"id":"https://openalex.org/C103057564","wikidata":"https://www.wikidata.org/wiki/Q4751139","display_name":"Analytic reasoning","level":3,"score":0.39879998564720154},{"id":"https://openalex.org/C89288958","wikidata":"https://www.wikidata.org/wiki/Q7301504","display_name":"Reasoning system","level":2,"score":0.35359999537467957},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3361999988555908},{"id":"https://openalex.org/C166052673","wikidata":"https://www.wikidata.org/wiki/Q83021","display_name":"Empirical evidence","level":2,"score":0.3208000063896179},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.31700000166893005},{"id":"https://openalex.org/C159032336","wikidata":"https://www.wikidata.org/wiki/Q2488768","display_name":"Non-monotonic logic","level":2,"score":0.2946999967098236},{"id":"https://openalex.org/C83725634","wikidata":"https://www.wikidata.org/wiki/Q7268699","display_name":"Qualitative reasoning","level":2,"score":0.2770000100135803},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.2696000039577484},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.2581999897956848},{"id":"https://openalex.org/C20162079","wikidata":"https://www.wikidata.org/wiki/Q1151406","display_name":"Case-based reasoning","level":2,"score":0.251800000667572}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.06787","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.06787","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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.2604.06787","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.06787","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"reasoning":[1,10,44,82,155],"models":[2,150],"(LRMs)":[3],"have":[4],"achieved":[5],"remarkable":[6],"performance":[7,161],"in":[8,27,74,110,170],"complex":[9],"tasks,":[11],"driven":[12],"by":[13,42,105,157],"their":[14],"powerful":[15],"inference-time":[16],"scaling":[17],"capability.":[18],"However,":[19],"LRMs":[20,171],"often":[21],"suffer":[22],"from":[23],"overthinking,":[24],"which":[25,117,132],"results":[26,146],"substantial":[28],"computational":[29],"redundancy":[30],"and":[31,64,96,127,172],"significantly":[32],"reduces":[33,154],"efficiency.":[34],"Early-exit":[35],"methods":[36],"aim":[37],"to":[38,87,140],"mitigate":[39],"this":[40,67],"issue":[41],"terminating":[43],"once":[45],"sufficient":[46,139],"evidence":[47],"has":[48],"been":[49],"generated,":[50],"yet":[51],"existing":[52],"approaches":[53],"mostly":[54],"rely":[55],"on":[56,147],"handcrafted":[57],"or":[58],"empirical":[59],"indicators":[60],"that":[61,83,152],"are":[62],"unreliable":[63],"impractical.":[65],"In":[66],"work,":[68],"we":[69],"introduce":[70],"Dynamic":[71],"Thought":[72,129],"Sufficiency":[73,130],"Reasoning":[75],"(DTSR),":[76],"a":[77],"novel":[78],"framework":[79],"for":[80,101,124,178],"efficient":[81],"enables":[84],"the":[85,90,98,135,142,148],"model":[86],"dynamically":[88],"assess":[89],"sufficiency":[91],"of":[92],"its":[93],"chain-of-thought":[94],"(CoT)":[95],"determine":[97],"optimal":[99],"point":[100],"early":[102,125],"exit.":[103],"Inspired":[104],"human":[106],"metacognition,":[107],"DTSR":[108,153],"operates":[109],"two":[111],"stages:":[112],"(1)":[113],"Reflection":[114],"Signal":[115],"Monitoring,":[116],"identifies":[118],"reflection":[119],"signals":[120],"as":[121],"potential":[122],"cues":[123],"exit,":[126],"(2)":[128],"Check,":[131],"evaluates":[133],"whether":[134],"current":[136],"CoT":[137],"is":[138],"derive":[141],"final":[143],"answer.":[144],"Experimental":[145],"Qwen3":[149],"show":[151],"length":[156],"28.9%-34.9%":[158],"with":[159],"minimal":[160],"loss,":[162],"effectively":[163],"mitigating":[164],"overthinking.":[165],"We":[166],"further":[167],"discuss":[168],"overconfidence":[169],"self-evaluation":[173],"paradigms,":[174],"providing":[175],"valuable":[176],"insights":[177],"early-exit":[179],"reasoning.":[180]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-04-10T00:00:00"}