{"id":"https://openalex.org/W4412888524","doi":"https://doi.org/10.18653/v1/2025.findings-acl.362","title":"Task Calibration: Calibrating Large Language Models on Inference Tasks","display_name":"Task Calibration: Calibrating Large Language Models on Inference Tasks","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412888524","doi":"https://doi.org/10.18653/v1/2025.findings-acl.362"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.findings-acl.362","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.362","pdf_url":"https://aclanthology.org/2025.findings-acl.362.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":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.findings-acl.362.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100425080","display_name":"Yingjie Li","orcid":"https://orcid.org/0000-0003-4015-4576"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yingjie Li","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101308917","display_name":"Yun Luo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yun Luo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101690037","display_name":"Xiaotian Xie","orcid":"https://orcid.org/0000-0002-4080-6901"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiaotian Xie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5100333738","display_name":"Yue Zhang","orcid":"https://orcid.org/0000-0002-6327-5023"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yue Zhang","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":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"6937","last_page":"6951"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.98580002784729,"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":0.98580002784729,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.9577000141143799,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7955451011657715},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.7040590047836304},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.703690230846405},{"id":"https://openalex.org/keywords/calibration","display_name":"Calibration","score":0.6897773742675781},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.5049849152565002},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4748245179653168},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.45179250836372375},{"id":"https://openalex.org/keywords/machine-learning","display_name":"Machine learning","score":0.36644184589385986},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.10437169671058655},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.09936222434043884}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7955451011657715},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.7040590047836304},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.703690230846405},{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.6897773742675781},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5049849152565002},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4748245179653168},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.45179250836372375},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.36644184589385986},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.10437169671058655},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.09936222434043884},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.findings-acl.362","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.362","pdf_url":"https://aclanthology.org/2025.findings-acl.362.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"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.findings-acl.362","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.findings-acl.362","pdf_url":"https://aclanthology.org/2025.findings-acl.362.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":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412888524.pdf","grobid_xml":"https://content.openalex.org/works/W4412888524.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2055243143","https://openalex.org/W1986418932","https://openalex.org/W2357796999","https://openalex.org/W4321636575","https://openalex.org/W2741131631","https://openalex.org/W2045526782","https://openalex.org/W2156919374","https://openalex.org/W1483472507","https://openalex.org/W1984019423","https://openalex.org/W3098003361"],"abstract_inverted_index":{"Large":[0],"language":[1,34,40,142],"models":[2],"(LLMs)":[3],"have":[4],"exhibited":[5],"impressive":[6],"zero-shot":[7,74,128],"performance":[8,66,86],"on":[9,32,50,97,107,122],"inference":[10,41],"tasks.However,":[11],"LLMs":[12,44,93],"may":[13,45,62],"suffer":[14],"from":[15],"spurious":[16],"correlations":[17],"between":[18],"input":[19],"texts":[20],"and":[21,75,100,139,155],"output":[22],"labels,":[23],"which":[24,83],"limits":[25],"LLMs'":[26],"ability":[27],"to":[28,64,94,152,159,170],"reason":[29,95],"based":[30,49,96],"purely":[31],"general":[33],"understanding.For":[35],"example,":[36],"in":[37,126,136],"the":[38,104,127,132,157],"natural":[39,141],"(NLI)":[42],"task,":[43],"make":[46],"predictions":[47],"primarily":[48],"premise":[51,99,109],"or":[52,110],"hypothesis,":[53,101],"rather":[54],"than":[55],"both":[56,98],"components.To":[57],"address":[58],"this":[59],"problem":[60],"that":[61,116,147],"lead":[63],"unexpected":[65],"degradation,":[67],"we":[68],"propose":[69],"task":[70,88],"calibration":[71,77,164],"(TC),":[72],"a":[73,119],"inference-only":[76],"method":[78],"inspired":[79],"by":[80],"mutual":[81],"information":[82],"recovers":[84],"LLM":[85],"through":[87],"reformulation.In":[89],"NLI,":[90],"TC":[91,117,135,148],"encourages":[92],"while":[102],"mitigating":[103],"models'":[105],"overreliance":[106],"individual":[108],"hypothesis":[111],"for":[112],"inference.Experimental":[113],"results":[114],"show":[115],"achieves":[118],"substantial":[120],"improvement":[121],"13":[123],"different":[124],"benchmarks":[125],"setup.We":[129],"further":[130],"validate":[131],"effectiveness":[133],"of":[134],"fewshot":[137],"setups":[138],"various":[140],"understanding":[143],"tasks.Further":[144],"analysis":[145],"indicates":[146],"is":[149],"also":[150],"robust":[151],"prompt":[153],"templates":[154],"has":[156],"potential":[158],"be":[160],"integrated":[161],"with":[162],"other":[163],"methods.We":[165],"publicly":[166],"release":[167],"our":[168],"code":[169],"facilitate":[171],"future":[172],"research":[173],"1":[174],".":[175]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}