{"id":"https://openalex.org/W4412886674","doi":"https://doi.org/10.18653/v1/2025.acl-long.188","title":"Influences on LLM Calibration: A Study of Response Agreement, Loss Functions, and Prompt Styles","display_name":"Influences on LLM Calibration: A Study of Response Agreement, Loss Functions, and Prompt Styles","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412886674","doi":"https://doi.org/10.18653/v1/2025.acl-long.188"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-long.188","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.188","pdf_url":"https://aclanthology.org/2025.acl-long.188.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":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.acl-long.188.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Yuxi Xia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuxi Xia","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063134162","display_name":"Pedro Araujo","orcid":"https://orcid.org/0000-0002-1962-6084"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pedro Henrique Luz De Araujo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011614450","display_name":"Klim Zaporojets","orcid":"https://orcid.org/0000-0003-4988-978X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Klim Zaporojets","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5046895021","display_name":"Benjamin Roth","orcid":"https://orcid.org/0000-0003-0362-0267"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Benjamin Roth","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":6,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"3740","last_page":"3761"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12560","display_name":"Nuclear Engineering Thermal-Hydraulics","score":0.6751999855041504,"subfield":{"id":"https://openalex.org/subfields/2202","display_name":"Aerospace Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12560","display_name":"Nuclear Engineering Thermal-Hydraulics","score":0.6751999855041504,"subfield":{"id":"https://openalex.org/subfields/2202","display_name":"Aerospace Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/calibration","display_name":"Calibration","score":0.6624497771263123},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.39408642053604126},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.24208033084869385},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.2302139401435852}],"concepts":[{"id":"https://openalex.org/C165838908","wikidata":"https://www.wikidata.org/wiki/Q736777","display_name":"Calibration","level":2,"score":0.6624497771263123},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.39408642053604126},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.24208033084869385},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2302139401435852}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.18653/v1/2025.acl-long.188","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.188","pdf_url":"https://aclanthology.org/2025.acl-long.188.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"},{"id":"pmh:oai:pure.atira.dk:openaire/326491de-0319-42c5-a43d-25ffc3a6c82c","is_oa":true,"landing_page_url":"https://pure.au.dk/portal/en/publications/326491de-0319-42c5-a43d-25ffc3a6c82c","pdf_url":null,"source":null,"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/conferenceObject"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.acl-long.188","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.188","pdf_url":"https://aclanthology.org/2025.acl-long.188.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":[{"id":"https://openalex.org/G5514542193","display_name":null,"funder_award_id":"10.47379/VRG19008","funder_id":"https://openalex.org/F4320321003","funder_display_name":"Vienna Science and Technology Fund"}],"funders":[{"id":"https://openalex.org/F4320320300","display_name":"European Commission","ror":"https://ror.org/00k4n6c32"},{"id":"https://openalex.org/F4320321003","display_name":"Vienna Science and Technology Fund","ror":"https://ror.org/01f9mc681"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412886674.pdf","grobid_xml":"https://content.openalex.org/works/W4412886674.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":{"Calibration,":[0],"the":[1,12,24,57,128],"alignment":[2],"between":[3],"model":[4,82],"confidence":[5,84],"and":[6,33,49,63,101,116,135,149],"prediction":[7],"accuracy,":[8],"is":[9],"critical":[10],"for":[11,83,131],"reliable":[13],"deployment":[14],"of":[15,26,36,60],"large":[16],"language":[17],"models":[18,137],"(LLMs).Existing":[19],"works":[20],"neglect":[21],"to":[22,29,92],"measure":[23],"generalization":[25],"their":[27],"methods":[28],"other":[30],"prompt":[31,51],"styles":[32],"different":[34],"sizes":[35],"LLMs.To":[37],"address":[38],"this,":[39],"we":[40,72,98],"define":[41],"a":[42,75],"controlled":[43],"experimental":[44],"setting":[45],"covering":[46],"12":[47],"LLMs":[48,62,91],"four":[50,109],"styles.We":[52],"additionally":[53],"investigate":[54],"if":[55],"incorporating":[56],"response":[58,114],"agreement":[59,115],"multiple":[61,90],"an":[64,80],"appropriate":[65],"loss":[66,118],"function":[67],"can":[68],"improve":[69,119],"calibration":[70,120,140],"performance.Concretely,":[71],"build":[73],"Calib-n,":[74],"novel":[76],"framework":[77],"that":[78,86,112,124],"trains":[79],"auxiliary":[81,132,136],"estimation":[85],"aggregates":[87],"responses":[88],"from":[89,121],"capture":[93],"inter-model":[94],"agreement.To":[95],"optimize":[96],"calibration,":[97],"integrate":[99],"focal":[100,117],"AUC":[102],"surrogate":[103],"losses":[104],"alongside":[105],"binary":[106],"cross-entropy.Experiments":[107],"across":[108,142],"datasets":[110],"demonstrate":[111,138],"both":[113],"baselines.We":[122],"find":[123],"few-shot":[125],"prompts":[126],"are":[127],"most":[129],"effective":[130],"model-based":[133],"methods,":[134],"robust":[139],"performance":[141],"accuracy":[143],"variations,":[144],"outperforming":[145],"LLMs'":[146],"internal":[147],"probabilities":[148],"verbalized":[150],"confidences.":[151],"1":[152],"Sigmoid":[153],"...":[154]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":3}],"updated_date":"2026-07-19T07:52:34.831488","created_date":"2025-10-10T00:00:00"}