{"id":"https://openalex.org/W4416036601","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.348","title":"CoCoA: Confidence- and Context-Aware Adaptive Decoding for Resolving Knowledge Conflicts in Large Language Models","display_name":"CoCoA: Confidence- and Context-Aware Adaptive Decoding for Resolving Knowledge Conflicts in Large Language Models","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416036601","doi":"https://doi.org/10.18653/v1/2025.emnlp-main.348"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.emnlp-main.348","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.348","pdf_url":"https://aclanthology.org/2025.emnlp-main.348.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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.emnlp-main.348.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101661660","display_name":"Ashish Khandelwal","orcid":"https://orcid.org/0000-0003-0903-5579"},"institutions":[{"id":"https://openalex.org/I4210124949","display_name":"Microsoft Research (India)","ror":"https://ror.org/02w7f3w92","country_code":"IN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210124949"]},{"id":"https://openalex.org/I4210162141","display_name":"Microsoft (India)","ror":"https://ror.org/04ww0w091","country_code":"IN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210162141"]}],"countries":["IN"],"is_corresponding":true,"raw_author_name":"Anant Khandelwal","raw_affiliation_strings":["Microsoft , India"],"affiliations":[{"raw_affiliation_string":"Microsoft , India","institution_ids":["https://openalex.org/I4210162141","https://openalex.org/I4210124949"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070111727","display_name":"Manish K. Gupta","orcid":"https://orcid.org/0000-0003-4226-0364"},"institutions":[{"id":"https://openalex.org/I4210124949","display_name":"Microsoft Research (India)","ror":"https://ror.org/02w7f3w92","country_code":"IN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210124949"]},{"id":"https://openalex.org/I4210162141","display_name":"Microsoft (India)","ror":"https://ror.org/04ww0w091","country_code":"IN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210162141"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Manish Gupta","raw_affiliation_strings":["Microsoft , India"],"affiliations":[{"raw_affiliation_string":"Microsoft , India","institution_ids":["https://openalex.org/I4210162141","https://openalex.org/I4210124949"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5064683413","display_name":"Puneet Agrawal","orcid":"https://orcid.org/0000-0002-5745-7330"},"institutions":[{"id":"https://openalex.org/I4210124949","display_name":"Microsoft Research (India)","ror":"https://ror.org/02w7f3w92","country_code":"IN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210124949"]},{"id":"https://openalex.org/I4210162141","display_name":"Microsoft (India)","ror":"https://ror.org/04ww0w091","country_code":"IN","type":"company","lineage":["https://openalex.org/I1290206253","https://openalex.org/I4210162141"]}],"countries":["IN"],"is_corresponding":false,"raw_author_name":"Puneet Agrawal","raw_affiliation_strings":["Microsoft , India"],"affiliations":[{"raw_affiliation_string":"Microsoft , India","institution_ids":["https://openalex.org/I4210162141","https://openalex.org/I4210124949"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5101661660"],"corresponding_institution_ids":["https://openalex.org/I4210124949","https://openalex.org/I4210162141"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.18457326,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"6846","last_page":"6866"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.44839999079704285,"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/T10028","display_name":"Topic Modeling","score":0.44839999079704285,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.15960000455379486,"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":0.049300000071525574,"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/decoding-methods","display_name":"Decoding methods","score":0.5026000142097473},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.3614000082015991},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.3393000066280365},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.2872999906539917},{"id":"https://openalex.org/keywords/knowledge-based-systems","display_name":"Knowledge-based systems","score":0.28040000796318054},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.2669999897480011}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.673799991607666},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.5026000142097473},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.474700003862381},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.46970000863075256},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3614000082015991},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.3393000066280365},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.30410000681877136},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2872999906539917},{"id":"https://openalex.org/C115925183","wikidata":"https://www.wikidata.org/wiki/Q1412694","display_name":"Knowledge-based systems","level":2,"score":0.28040000796318054},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.2669999897480011},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.2669000029563904},{"id":"https://openalex.org/C161301231","wikidata":"https://www.wikidata.org/wiki/Q3478658","display_name":"Knowledge representation and reasoning","level":2,"score":0.26570001244544983},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.26420000195503235},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.25380000472068787},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.25220000743865967}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.emnlp-main.348","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.348","pdf_url":"https://aclanthology.org/2025.emnlp-main.348.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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.emnlp-main.348","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.emnlp-main.348","pdf_url":"https://aclanthology.org/2025.emnlp-main.348.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 2025 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416036601.pdf","grobid_xml":"https://content.openalex.org/works/W4416036601.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Faithful":[0],"generation":[1],"in":[2,33,80,112,130,168,192,206],"large":[3],"language":[4],"models":[5],"(LLMs)":[6],"is":[7],"challenged":[8],"by":[9,56,134],"knowledge":[10],"conflicts":[11],"between":[12,69],"parametric":[13,71],"memory":[14],"and":[15,29,51,62,65,72,94,127,132,155],"external":[16],"context.Existing":[17],"contrastive":[18],"decoding":[19],"methods":[20],"tuned":[21],"specifically":[22],"to":[23,116,122,136,148],"handle":[24],"conflict":[25,35,49,55,82,149],"often":[26],"lack":[27],"adaptability":[28],"can":[30],"degrade":[31],"performance":[32,78,103],"low":[34,81],"settings.We":[36],"introduce":[37],"COCOA":[38,75],"(Confidence-and":[39],"Context-Aware":[40],"Adaptive":[41],"Decoding),":[42],"a":[43],"novel":[44],"token-level":[45],"algorithm":[46],"for":[47],"principled":[48],"resolution":[50],"enhanced":[52],"faithfulness.COCOA":[53],"resolves":[54],"utilizing":[57],"confidence-aware":[58],"measures":[59],"(entropy":[60],"gap":[61],"contextual":[63,73],"peakedness)":[64],"the":[66,70,123,184,189,203],"generalized":[67],"divergence":[68],"distributions.Crucially,":[74],"maintains":[76],"strong":[77,105,124],"even":[79],"settings.Extensive":[83],"experiments":[84],"across":[85,141],"multiple":[86,166],"LLMs":[87],"on":[88,119,139,179],"diverse":[89],"Question":[90,96],"Answering":[91,97],"(QA),":[92],"Summarization,":[93],"Long-Form":[95],"(LFQA)":[98],"benchmarks":[99],"demonstrate":[100],"CO-COA's":[101],"state-of-the-art":[102],"over":[104],"baselines":[106],"like":[107],"ADACAD.It":[108],"yields":[109],"significant":[110],"gains":[111],"QA":[113],"accuracy,":[114],"up":[115,135],"9.2":[117],"points":[118,138],"average":[120,140],"compared":[121],"baseline":[125],"ADACAD,":[126],"improves":[128],"factuality":[129],"summarization":[131],"LFQA":[133],"2.5":[137],"key":[142],"benchmarks.Additionally,":[143],"it":[144],"demonstrates":[145],"superior":[146],"sensitivity":[147],"variations.CO-COA":[150],"enables":[151],"more":[152,157],"informed,":[153],"context-aware,":[154],"ultimately":[156],"faithful":[158],"token":[159],"generation.High":[160],"Conflict":[161],"Context":[162],"Tom":[163],"Hanks":[164],"played":[165,183,188,202],"roles":[167],"Polar":[169,193,207],"Express,":[170],"including":[171],"Conductor,":[172],"Hero":[173],"Boy...":[174],"Dylan":[175],"Cash":[176],"as":[177],"Boy":[178,205],"Train":[180],"Dante":[181],"Pastula":[182],"Lile":[185],"Boy.":[186],"QuestionWho":[187],"little":[190],"boy":[191],"Express?":[194],"Parametric":[195],"Knowledge":[196],"(LM)The":[197],"LM":[198],"believes":[199],"Daryl":[200],"Sabara":[201],"Little":[204],"Express.":[208]},"counts_by_year":[],"updated_date":"2026-03-11T06:11:40.159057","created_date":"2025-11-08T00:00:00"}