{"id":"https://openalex.org/W4416034517","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.1362","title":"Quantifying Uncertainty in Natural Language Explanations of Large Language Models for Question Answering","display_name":"Quantifying Uncertainty in Natural Language Explanations of Large Language Models for Question Answering","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4416034517","doi":"https://doi.org/10.18653/v1/2025.findings-emnlp.1362"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.1362","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.1362","pdf_url":"https://aclanthology.org/2025.findings-emnlp.1362.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":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-emnlp.1362.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101806898","display_name":"Yangyi Li","orcid":"https://orcid.org/0009-0008-6564-1710"},"institutions":[{"id":"https://openalex.org/I173911158","display_name":"Iowa State University","ror":"https://ror.org/04rswrd78","country_code":"US","type":"education","lineage":["https://openalex.org/I173911158"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yangyi Li","raw_affiliation_strings":["Department of Computer Science , Iowa State University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Iowa State University","institution_ids":["https://openalex.org/I173911158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5016035883","display_name":"Mengdi Huai","orcid":"https://orcid.org/0000-0001-6368-5973"},"institutions":[{"id":"https://openalex.org/I173911158","display_name":"Iowa State University","ror":"https://ror.org/04rswrd78","country_code":"US","type":"education","lineage":["https://openalex.org/I173911158"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mengdi Huai","raw_affiliation_strings":["Department of Computer Science , Iowa State University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Iowa State University","institution_ids":["https://openalex.org/I173911158"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I173911158"],"apc_list":null,"apc_paid":null,"fwci":3.3016,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":{"value":0.93295564,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"25005","last_page":"25013"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.5550000071525574,"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":0.5550000071525574,"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/T10028","display_name":"Topic Modeling","score":0.08699999749660492,"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":0.06430000066757202,"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/question-answering","display_name":"Question answering","score":0.8021000027656555},{"id":"https://openalex.org/keywords/natural-language","display_name":"Natural language","score":0.5454000234603882},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4034999907016754},{"id":"https://openalex.org/keywords/language-identification","display_name":"Language identification","score":0.3718000054359436},{"id":"https://openalex.org/keywords/natural","display_name":"Natural (archaeology)","score":0.34709998965263367},{"id":"https://openalex.org/keywords/computational-linguistics","display_name":"Computational linguistics","score":0.31200000643730164}],"concepts":[{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.8021000027656555},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6067000031471252},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5709999799728394},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.5454000234603882},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5062999725341797},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4034999907016754},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.3779999911785126},{"id":"https://openalex.org/C129792486","wikidata":"https://www.wikidata.org/wiki/Q1050419","display_name":"Language identification","level":3,"score":0.3718000054359436},{"id":"https://openalex.org/C2776608160","wikidata":"https://www.wikidata.org/wiki/Q4785462","display_name":"Natural (archaeology)","level":2,"score":0.34709998965263367},{"id":"https://openalex.org/C155092808","wikidata":"https://www.wikidata.org/wiki/Q182557","display_name":"Computational linguistics","level":2,"score":0.31200000643730164},{"id":"https://openalex.org/C83479923","wikidata":"https://www.wikidata.org/wiki/Q2063748","display_name":"Universal Networking Language","level":4,"score":0.2881999909877777},{"id":"https://openalex.org/C2779439875","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Natural language understanding","level":3,"score":0.27489998936653137},{"id":"https://openalex.org/C2779313563","wikidata":"https://www.wikidata.org/wiki/Q17072565","display_name":"On Language","level":2,"score":0.2502000033855438}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-emnlp.1362","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.1362","pdf_url":"https://aclanthology.org/2025.findings-emnlp.1362.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":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-emnlp.1362","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-emnlp.1362","pdf_url":"https://aclanthology.org/2025.findings-emnlp.1362.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":"Findings of the Association for Computational Linguistics: EMNLP 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G2453969620","display_name":"CAREER: Enabling Reliable Uncertainty-Aware Decision Making with Unreliable Data","funder_award_id":"2442750","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G4879695479","display_name":"Collaborative Research: SaTC: CORE: Small: Security and Privacy in Machine Unlearning","funder_award_id":"2350332","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4416034517.pdf","grobid_xml":"https://content.openalex.org/works/W4416034517.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1,34,40,88,107,146],"models":[2,65],"(LLMs)":[3],"have":[4],"shown":[5],"strong":[6],"capabilities,":[7],"enabling":[8],"concise,":[9],"contextaware":[10],"answers":[11],"in":[12,20,51,94,125,131,153],"question":[13],"answering":[14],"(QA)":[15],"tasks.The":[16],"lack":[17],"of":[18,59,118,123,184],"transparency":[19],"complex":[21],"LLMs":[22,50,119],"has":[23],"inspired":[24],"extensive":[25],"research":[26],"aimed":[27],"at":[28],"developing":[29],"methods":[30],"to":[31,45,48,79,113],"explain":[32,49],"large":[33],"behaviors.Among":[35],"existing":[36,75],"explanation":[37],"methods,":[38],"natural":[39,87,106,145],"explanations":[41,108],"stand":[42],"out":[43],"due":[44,112],"their":[46],"ability":[47],"a":[52,137,154,162],"selfexplanatory":[53],"manner":[54],"and":[55,120,156],"enable":[56],"the":[57,64,96,114,121,181],"understanding":[58,95],"model":[60],"behaviors":[61],"even":[62,173],"when":[63],"are":[66],"closed-source.However,":[67],"despite":[68],"these":[69,85,99,143],"promising":[70],"advancements,":[71],"there":[72],"is":[73,92,109],"no":[74],"work":[76],"studying":[77],"how":[78],"provide":[80],"valid":[81,102,150,170],"uncertainty":[82,90,103,139,151,165,171],"guarantees":[83,152,172],"for":[84,105,142],"generated":[86,144],"explanations.Such":[89],"quantification":[91],"critical":[93],"confidence":[97],"behind":[98],"explanations.Notably,":[100],"generating":[101],"estimates":[104],"particularly":[110],"challenging":[111],"auto-regressive":[115],"generation":[116],"process":[117],"presence":[122],"noise":[124],"medical":[126],"inquiries.To":[127],"bridge":[128],"this":[129,132],"gap,":[130],"work,":[133],"we":[134,159],"first":[135],"propose":[136],"novel":[138,163],"estimation":[140,166],"framework":[141],"explanations,":[147],"which":[148],"provides":[149],"post-hoc":[155],"model-agnostic":[157],"manner.Additionally,":[158],"also":[160],"design":[161],"robust":[164],"method":[167],"that":[168],"maintains":[169],"under":[174],"noise.Extensive":[175],"experiments":[176],"on":[177],"QA":[178],"tasks":[179],"demonstrate":[180],"desired":[182],"performance":[183],"our":[185],"methods.":[186]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-06-26T08:34:08.712188","created_date":"2025-11-08T00:00:00"}