{"id":"https://openalex.org/W4412888125","doi":"https://doi.org/10.18653/v1/2025.findings-acl.847","title":"Evaluating LLMs\u2019 Assessment of Mixed-Context Hallucination Through the Lens of Summarization","display_name":"Evaluating LLMs\u2019 Assessment of Mixed-Context Hallucination Through the Lens of Summarization","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412888125","doi":"https://doi.org/10.18653/v1/2025.findings-acl.847"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-acl.847","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.847","pdf_url":"https://aclanthology.org/2025.findings-acl.847.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: ACL 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-acl.847.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5005849122","display_name":"Siya Qi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Siya Qi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103013578","display_name":"Rui Cao","orcid":"https://orcid.org/0000-0001-9577-9667"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rui Cao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066124369","display_name":"Yuanyuan He","orcid":"https://orcid.org/0000-0003-1804-0302"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yulan He","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5017784727","display_name":"Zheng Yuan","orcid":"https://orcid.org/0000-0001-8788-6817"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng Yuan","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.14896047,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"16480","last_page":"16503"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11710","display_name":"Biomedical Text Mining and Ontologies","score":0.9387000203132629,"subfield":{"id":"https://openalex.org/subfields/1312","display_name":"Molecular Biology"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},"topics":[{"id":"https://openalex.org/T11710","display_name":"Biomedical Text Mining and Ontologies","score":0.9387000203132629,"subfield":{"id":"https://openalex.org/subfields/1312","display_name":"Molecular Biology"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/automatic-summarization","display_name":"Automatic summarization","score":0.8572447299957275},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.7088959217071533},{"id":"https://openalex.org/keywords/lens","display_name":"Lens (geology)","score":0.5873110294342041},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.49747946858406067},{"id":"https://openalex.org/keywords/geology","display_name":"Geology","score":0.2566099762916565},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2543204724788666},{"id":"https://openalex.org/keywords/paleontology","display_name":"Paleontology","score":0.08349448442459106}],"concepts":[{"id":"https://openalex.org/C170858558","wikidata":"https://www.wikidata.org/wiki/Q1394144","display_name":"Automatic summarization","level":2,"score":0.8572447299957275},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.7088959217071533},{"id":"https://openalex.org/C15336307","wikidata":"https://www.wikidata.org/wiki/Q1766051","display_name":"Lens (geology)","level":2,"score":0.5873110294342041},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.49747946858406067},{"id":"https://openalex.org/C127313418","wikidata":"https://www.wikidata.org/wiki/Q1069","display_name":"Geology","level":0,"score":0.2566099762916565},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2543204724788666},{"id":"https://openalex.org/C151730666","wikidata":"https://www.wikidata.org/wiki/Q7205","display_name":"Paleontology","level":1,"score":0.08349448442459106}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.18653/v1/2025.findings-acl.847","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.847","pdf_url":"https://aclanthology.org/2025.findings-acl.847.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: ACL 2025","raw_type":"proceedings-article"},{"id":"pmh:oai:kclpure.kcl.ac.uk:openaire/2aec0e5d-e3b6-4f3c-9ec0-5af823837a21","is_oa":true,"landing_page_url":"https://kclpure.kcl.ac.uk/portal/en/publications/2aec0e5d-e3b6-4f3c-9ec0-5af823837a21","pdf_url":null,"source":{"id":"https://openalex.org/S4306400216","display_name":"Research Portal (King's College London)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I183935753","host_organization_name":"King's College London","host_organization_lineage":["https://openalex.org/I183935753"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Qi, S, Cao, R, He, Y & Yuan, Z 2025, Evaluating LLMs\u2019 Assessment of Mixed-Context Hallucination Through the Lens of Summarization. in W Che, J Nabende, E Shutova & M T Pilehvar (eds), Findings of the Association for Computational Linguistics : ACL 2025. Vienna, pp. 16480\u201316503. https://doi.org/10.18653/v1/2025.findings-acl.847","raw_type":"info:eu-repo/semantics/publishedVersion"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-acl.847","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.847","pdf_url":"https://aclanthology.org/2025.findings-acl.847.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: ACL 2025","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G2577708131","display_name":"Turing AI Fellowship: Event-Centric Framework for Natural Language Understanding","funder_award_id":"EP/V020579/1","funder_id":"https://openalex.org/F4320334627","funder_display_name":"Engineering and Physical Sciences Research Council"},{"id":"https://openalex.org/G6382798043","display_name":"Turing AI Fellowship: Event-Centric Framework for Natural Language Understanding","funder_award_id":"EP/V020579/2","funder_id":"https://openalex.org/F4320334627","funder_display_name":"Engineering and Physical Sciences Research Council"}],"funders":[{"id":"https://openalex.org/F4320320285","display_name":"King's College London","ror":"https://ror.org/0220mzb33"},{"id":"https://openalex.org/F4320322725","display_name":"China Scholarship Council","ror":"https://ror.org/04atp4p48"},{"id":"https://openalex.org/F4320334627","display_name":"Engineering and Physical Sciences Research Council","ror":"https://ror.org/0439y7842"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412888125.pdf","grobid_xml":"https://content.openalex.org/works/W4412888125.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/W2366403280","https://openalex.org/W1495108544","https://openalex.org/W2091301346","https://openalex.org/W3148229873","https://openalex.org/W4389760904","https://openalex.org/W2150160875","https://openalex.org/W2351187795"],"abstract_inverted_index":{"With":[0],"the":[1,102,114],"rapid":[2],"development":[3],"of":[4,81,104],"large":[5],"language":[6],"models":[7,80],"(LLMs),":[8],"LLM-as-a-judge":[9],"has":[10],"emerged":[11],"as":[12,53],"a":[13,54,108],"widely":[14],"adopted":[15],"approach":[16],"for":[17,130],"text":[18],"quality":[19],"evaluation,":[20],"including":[21],"hallucination":[22,95,133],"evaluation.While":[23],"previous":[24],"studies":[25],"have":[26],"focused":[27],"exclusively":[28],"on":[29],"single-context":[30],"evaluation":[31],"(e.g.,":[32],"discourse":[33],"faithfulness":[34],"or":[35],"world":[36],"factuality),":[37],"realworld":[38],"hallucinations":[39],"typically":[40],"involve":[41],"mixed":[42],"contexts,":[43],"which":[44],"remains":[45],"inadequately":[46],"evaluated.In":[47],"this":[48],"study,":[49],"we":[50],"use":[51],"summarization":[52],"representative":[55],"task":[56],"to":[57],"comprehensively":[58],"evaluate":[59],"LLMs'":[60,88,124],"capability":[61],"in":[62,94,118],"detecting":[63],"mixedcontext":[64],"hallucinations,":[65,106],"specifically":[66],"distinguishing":[67],"between":[68,123],"factual":[69,105],"and":[70,78,112,127],"non-factual":[71],"hallucinations.Through":[72],"extensive":[73],"experiments":[74],"across":[75],"direct":[76],"generation":[77],"retrieval-based":[79],"varying":[82],"scales,":[83],"our":[84],"main":[85],"observations":[86],"are:(1)":[87],"intrinsic":[89,125],"knowledge":[90,120,126],"introduces":[91],"inherent":[92],"biases":[93,99],"evaluation;":[96],"(2)":[97],"these":[98],"particularly":[100],"impact":[101],"detection":[103],"yielding":[107],"significant":[109],"performance":[110],"bottleneck;":[111],"(3)":[113],"fundamental":[115],"challenge":[116],"lies":[117],"effective":[119],"utilization,":[121],"balancing":[122],"external":[128],"context":[129],"accurate":[131],"mixed-context":[132],"evaluation.":[134],"1":[135]},"counts_by_year":[],"updated_date":"2026-03-11T06:11:40.159057","created_date":"2025-10-10T00:00:00"}