{"id":"https://openalex.org/W4415980656","doi":"https://doi.org/10.48550/arxiv.2504.03579","title":"Hallucination Detection on a Budget: Efficient Bayesian Estimation of Semantic Entropy","display_name":"Hallucination Detection on a Budget: Efficient Bayesian Estimation of Semantic Entropy","publication_year":2025,"publication_date":"2025-04-04","ids":{"openalex":"https://openalex.org/W4415980656","doi":"https://doi.org/10.48550/arxiv.2504.03579"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2504.03579","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2504.03579","pdf_url":"https://arxiv.org/pdf/2504.03579","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2504.03579","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5043239539","display_name":"Kamil Ciosek","orcid":"https://orcid.org/0000-0002-0238-9393","observed_orcids":["https://orcid.org/0000-0002-0238-9393"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ciosek, Kamil","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047511363","display_name":"Nicol\u00f2 Felicioni","orcid":"https://orcid.org/0000-0002-3555-7760","observed_orcids":["https://orcid.org/0000-0002-3555-7760"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Felicioni, Nicol\u00f2","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5024789936","display_name":"Sina Ghiassian","orcid":"https://orcid.org/0000-0001-5853-5049","observed_orcids":["https://orcid.org/0000-0001-5853-5049"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ghiassian, Sina","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":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.8057000041007996,"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":0.8057000041007996,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.08829999715089798,"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":0.058400001376867294,"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/hallucination-detection","display_name":"hallucination detection","score":0.9229999780654907},{"id":"https://openalex.org/keywords/semantic-entropy","display_name":"semantic entropy","score":0.8970000147819519},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8040000200271606},{"id":"https://openalex.org/keywords/bayesian-estimation","display_name":"Bayesian estimation","score":0.9779999852180481},{"id":"https://openalex.org/keywords/adaptive-sampling","display_name":"adaptive sampling","score":0.6589999794960022}],"concepts":[{"id":"https://openalex.org/C185429906","wikidata":"https://www.wikidata.org/wiki/Q1130160","display_name":"Estimator","level":2,"score":0.6715999841690063},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.583899974822998},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.5758000016212463},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.571399986743927},{"id":"https://openalex.org/C106301342","wikidata":"https://www.wikidata.org/wiki/Q4117933","display_name":"Entropy (arrow of time)","level":2,"score":0.5498999953269958},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.41920000314712524},{"id":"https://openalex.org/C9679016","wikidata":"https://www.wikidata.org/wiki/Q1417473","display_name":"Principle of maximum entropy","level":2,"score":0.3684999942779541},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.3659999966621399},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3637999892234802},{"id":"https://openalex.org/C171752962","wikidata":"https://www.wikidata.org/wiki/Q255166","display_name":"Kullback\u2013Leibler divergence","level":2,"score":0.34220001101493835},{"id":"https://openalex.org/C68022304","wikidata":"https://www.wikidata.org/wiki/Q842217","display_name":"Bayes estimator","level":3,"score":0.34060001373291016},{"id":"https://openalex.org/C198531522","wikidata":"https://www.wikidata.org/wiki/Q485146","display_name":"Sample (material)","level":2,"score":0.33660000562667847},{"id":"https://openalex.org/C149441793","wikidata":"https://www.wikidata.org/wiki/Q200726","display_name":"Probability distribution","level":2,"score":0.3325999975204468},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.29679998755455017},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2793999910354614},{"id":"https://openalex.org/C3020402766","wikidata":"https://www.wikidata.org/wiki/Q104376712","display_name":"Prior information","level":2,"score":0.27570000290870667},{"id":"https://openalex.org/C66696666","wikidata":"https://www.wikidata.org/wiki/Q17105612","display_name":"Sample entropy","level":3,"score":0.2696000039577484},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.26499998569488525},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.25870001316070557},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2535000145435333},{"id":"https://openalex.org/C49781872","wikidata":"https://www.wikidata.org/wiki/Q1045555","display_name":"Maximum likelihood","level":2,"score":0.25040000677108765}],"mesh":[],"locations_count":3,"locations":[{"id":"pmh:oai:arXiv.org:2504.03579","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2504.03579","pdf_url":"https://arxiv.org/pdf/2504.03579","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:oai:re.public.polimi.it:11311/1297627","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2504.03579v2","pdf_url":null,"source":{"id":"https://openalex.org/S4306400312","display_name":"Virtual Community of Pathological Anatomy (University of Castilla La Mancha)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I79189158","host_organization_name":"University of Castilla-La Mancha","host_organization_lineage":["https://openalex.org/I79189158"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/article"},{"id":"doi:10.48550/arxiv.2504.03579","doi":"https://doi.org/10.48550/arxiv.2504.03579","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2504.03579","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2504.03579","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2504.03579","pdf_url":"https://arxiv.org/pdf/2504.03579","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4415980656.pdf","grobid_xml":"https://content.openalex.org/works/W4415980656.grobid-xml"},"referenced_works_count":15,"referenced_works":["https://openalex.org/W1858602683","https://openalex.org/W2612431505","https://openalex.org/W2947497897","https://openalex.org/W2963748441","https://openalex.org/W3021338988","https://openalex.org/W3093852412","https://openalex.org/W3138392969","https://openalex.org/W4285429195","https://openalex.org/W4309674289","https://openalex.org/W4321594373","https://openalex.org/W4394972078","https://openalex.org/W4400023100","https://openalex.org/W4405655726","https://openalex.org/W4415799082","https://openalex.org/W4415800481"],"related_works":["https://openalex.org/W7140291657","https://openalex.org/W7221509191","https://openalex.org/W4400023100","https://openalex.org/W7161725205","https://openalex.org/W4415239845","https://openalex.org/W7162650089","https://openalex.org/W7169168860","https://openalex.org/W4399803256","https://openalex.org/W4389524154","https://openalex.org/W4407244904"],"abstract_inverted_index":{"Detecting":[0],"whether":[1],"an":[2,6],"LLM":[3],"hallucinates":[4],"is":[5,16,127],"important":[7],"research":[8],"challenge.":[9],"One":[10],"promising":[11],"way":[12],"of":[13,26,29,58,66,79,102,115],"doing":[14,38],"so":[15,82],"to":[17,46,75,110],"estimate":[18],"the":[19,27,49,69,77,97,112,135],"semantic":[20,59],"entropy":[21,60],"(Farquhar":[22],"et":[23,107],"al.,":[24],"2024)":[25],"distribution":[28],"generated":[30],"sequences.":[31],"We":[32,89],"propose":[33],"a":[34,54,63],"new":[35],"algorithm":[36],"for":[37,62],"that,":[39],"with":[40,130],"two":[41],"main":[42],"advantages.":[43],"First,":[44],"due":[45],"us":[47],"taking":[48],"Bayesian":[50],"approach,":[51],"we":[52,72],"achieve":[53,111],"much":[55],"better":[56],"quality":[57,114],"estimates":[61],"given":[64],"budget":[65],"samples":[67,80,103],"from":[68,134],"LLM.":[70,136],"Second,":[71],"are":[73],"able":[74],"tune":[76],"number":[78],"adaptively":[81],"that":[83,92],"`harder'":[84],"contexts":[85],"receive":[86],"more":[87],"samples.":[88],"demonstrate":[90],"empirically":[91],"our":[93,125],"approach":[94],"systematically":[95],"beats":[96],"baselines,":[98],"requiring":[99],"only":[100],"53%":[101],"used":[104],"by":[105,120],"Farquhar":[106],"al.":[108],"(2024)":[109],"same":[113],"hallucination":[116],"detection":[117],"as":[118],"measured":[119],"AUROC.":[121],"Moreover,":[122],"quite":[123],"counterintuitively,":[124],"estimator":[126],"useful":[128],"even":[129],"just":[131],"one":[132],"sample":[133]},"counts_by_year":[{"year":2026,"cited_by_count":6}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2025-10-10T00:00:00"}