{"id":"https://openalex.org/W4412886705","doi":"https://doi.org/10.18653/v1/2025.acl-long.203","title":"Confidence v.s. Critique: A Decomposition of Self-Correction Capability for LLMs","display_name":"Confidence v.s. Critique: A Decomposition of Self-Correction Capability for LLMs","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412886705","doi":"https://doi.org/10.18653/v1/2025.acl-long.203"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-long.203","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.203","pdf_url":"https://aclanthology.org/2025.acl-long.203.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.acl-long.203.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101961856","display_name":"Zhe Yang","orcid":"https://orcid.org/0000-0003-4051-6610"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhe Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019898400","display_name":"Yuhan Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yichang Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100355030","display_name":"Yudong Wang","orcid":"https://orcid.org/0000-0001-8302-5518"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yudong Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068791753","display_name":"Zhao Xu","orcid":"https://orcid.org/0000-0003-4480-7394"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ziyao Xu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100612233","display_name":"Junyang Lin","orcid":"https://orcid.org/0000-0001-9931-383X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Junyang Lin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5110285832","display_name":"Zhifang Sui","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhifang Sui","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":23.248,"has_fulltext":true,"cited_by_count":4,"citation_normalized_percentile":{"value":0.99196723,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"3998","last_page":"4014"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13075","display_name":"Legal Systems and Judicial Processes","score":0.9829000234603882,"subfield":{"id":"https://openalex.org/subfields/3320","display_name":"Political Science and International Relations"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T13075","display_name":"Legal Systems and Judicial Processes","score":0.9829000234603882,"subfield":{"id":"https://openalex.org/subfields/3320","display_name":"Political Science and International Relations"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T14030","display_name":"Taxation and Legal Issues","score":0.968500018119812,"subfield":{"id":"https://openalex.org/subfields/1402","display_name":"Accounting"},"field":{"id":"https://openalex.org/fields/14","display_name":"Business, Management and Accounting"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T13972","display_name":"Business Law and Ethics","score":0.965499997138977,"subfield":{"id":"https://openalex.org/subfields/1402","display_name":"Accounting"},"field":{"id":"https://openalex.org/fields/14","display_name":"Business, Management and Accounting"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/decomposition","display_name":"Decomposition","score":0.5856115221977234},{"id":"https://openalex.org/keywords/confidence-interval","display_name":"Confidence interval","score":0.4830551743507385},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.39303961396217346},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.27498215436935425},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.21770310401916504},{"id":"https://openalex.org/keywords/chemistry","display_name":"Chemistry","score":0.11513414978981018}],"concepts":[{"id":"https://openalex.org/C124681953","wikidata":"https://www.wikidata.org/wiki/Q339062","display_name":"Decomposition","level":2,"score":0.5856115221977234},{"id":"https://openalex.org/C44249647","wikidata":"https://www.wikidata.org/wiki/Q208498","display_name":"Confidence interval","level":2,"score":0.4830551743507385},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.39303961396217346},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.27498215436935425},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.21770310401916504},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.11513414978981018},{"id":"https://openalex.org/C178790620","wikidata":"https://www.wikidata.org/wiki/Q11351","display_name":"Organic chemistry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.acl-long.203","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.203","pdf_url":"https://aclanthology.org/2025.acl-long.203.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.acl-long.203","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.203","pdf_url":"https://aclanthology.org/2025.acl-long.203.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412886705.pdf","grobid_xml":"https://content.openalex.org/works/W4412886705.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"can":[4,108,131],"correct":[5,56],"their":[6],"self-generated":[7],"responses,":[8],"but":[9],"a":[10,20,71,134,151],"decline":[11,135],"in":[12,136,173],"accuracy":[13,180],"after":[14,44,181],"self-correction":[15,33,49,85,142,158],"is":[16,184],"also":[17,121],"witnessed.To":[18],"have":[19],"deeper":[21],"understanding":[22],"of":[23,35],"selfcorrection,":[24],"we":[25,46,94,104,149],"endeavor":[26],"to":[27,55,63,74,133,156],"decompose,":[28],"evaluate,":[29],"and":[30,38,43,58,66,91,98,167,176],"analyze":[31],"the":[32,48,123,126,137],"behaviors":[34],"LLMs.By":[36],"enumerating":[37],"analyzing":[39],"answer":[40],"correctness":[41],"before":[42],"self-correction,":[45],"decompose":[47],"capability":[50,86,159],"into":[51],"confidence":[52],"(being":[53],"confident":[54,115],"answers)":[57],"critique":[59],"(turning":[60],"wrong":[61],"answers":[62],"correct)":[64],"capabilities,":[65,78],"propose":[67],"two":[68,127],"metrics":[69],"from":[70],"probabilistic":[72],"perspective":[73],"measure":[75],"these":[76],"2":[77],"along":[79],"with":[80],"another":[81],"metric":[82],"for":[83],"overall":[84],"evaluation.Based":[87],"on":[88,187],"our":[89,168],"decomposition":[90],"evaluation":[92],"metrics,":[93],"conduct":[95],"extensive":[96],"experiments":[97],"draw":[99],"some":[100,112],"empirical":[101],"conclusions.For":[102],"example,":[103],"find":[105,122,150],"different":[106],"models":[107,113],"exhibit":[109],"distinct":[110],"behaviors:":[111],"are":[114,118],"while":[116],"others":[117],"more":[119],"critical.We":[120],"trade-off":[124],"between":[125],"capabilities":[128,175],"(i.e.improving":[129],"one":[130],"lead":[132],"other)":[138],"when":[139],"manipulating":[140],"model":[141],"behavior":[143],"by":[144,160],"prompts":[145],"or":[146],"in-context":[147],"learning.Further,":[148],"simple":[152],"yet":[153],"efficient":[154],"strategy":[155,169],"improve":[157],"transforming":[161],"Supervision":[162],"Fine-Tuning":[163],"(SFT)":[164],"data":[165],"format,":[166],"outperforms":[170],"vanilla":[171],"SFT":[172],"both":[174],"achieves":[177],"much":[178],"higher":[179],"self-correction.Our":[182],"code":[183],"publicly":[185],"available":[186],"GitHub.":[188]},"counts_by_year":[{"year":2026,"cited_by_count":4}],"updated_date":"2026-06-27T08:28:00.272161","created_date":"2025-10-10T00:00:00"}
