{"id":"https://openalex.org/W7160848286","doi":"https://doi.org/10.48550/arxiv.2605.06915","title":"LLMs are not (consistently) Bayesian: Quantifying internal (in)consistencies of LLMs' probabilistic beliefs","display_name":"LLMs are not (consistently) Bayesian: Quantifying internal (in)consistencies of LLMs' probabilistic beliefs","publication_year":2026,"publication_date":"2026-05-07","ids":{"openalex":"https://openalex.org/W7160848286","doi":"https://doi.org/10.48550/arxiv.2605.06915"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.06915","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.06915","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.06915","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5028711697","display_name":"Chacha Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Chacha","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002428461","display_name":"Matthew J\u00f6rke","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"J\u00f6rke, Matthew","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135904276","display_name":"Adam Goli\u0144ski","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Goli\u0144ski, Adam","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000858764","display_name":"Masha Fedzechkina","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fedzechkina, Masha","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135872294","display_name":"Guillermo Sapiro","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sapiro, Guillermo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130325089","display_name":"Sinead Williamson","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Williamson, Sinead","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5104684719","display_name":"Nicholas Foti","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Foti, Nicholas","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":7,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.36629998683929443,"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":0.36629998683929443,"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":0.18310000002384186,"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/T11303","display_name":"Bayesian Modeling and Causal Inference","score":0.06040000170469284,"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/probabilistic-logic","display_name":"Probabilistic logic","score":0.7396000027656555},{"id":"https://openalex.org/keywords/heuristic","display_name":"Heuristic","score":0.6383000016212463},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.6116999983787537},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.535099983215332},{"id":"https://openalex.org/keywords/measure","display_name":"Measure (data warehouse)","score":0.39879998564720154},{"id":"https://openalex.org/keywords/bayesian-inference","display_name":"Bayesian inference","score":0.3919000029563904},{"id":"https://openalex.org/keywords/information-processing","display_name":"Information processing","score":0.3783999979496002}],"concepts":[{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.7396000027656555},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.6383000016212463},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.6116999983787537},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.535099983215332},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5236999988555908},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4661000072956085},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4657000005245209},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.39879998564720154},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.3919000029563904},{"id":"https://openalex.org/C87868495","wikidata":"https://www.wikidata.org/wiki/Q750843","display_name":"Information processing","level":2,"score":0.3783999979496002},{"id":"https://openalex.org/C112930515","wikidata":"https://www.wikidata.org/wiki/Q4389547","display_name":"Risk analysis (engineering)","level":1,"score":0.3756999969482422},{"id":"https://openalex.org/C33724603","wikidata":"https://www.wikidata.org/wiki/Q812540","display_name":"Bayesian network","level":2,"score":0.32820001244544983},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.32330000400543213},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.32179999351501465},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.3154999911785126},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.28870001435279846},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.2858000099658966},{"id":"https://openalex.org/C207201462","wikidata":"https://www.wikidata.org/wiki/Q182505","display_name":"Bayes' theorem","level":3,"score":0.26980000734329224},{"id":"https://openalex.org/C113336015","wikidata":"https://www.wikidata.org/wiki/Q574010","display_name":"Complete information","level":2,"score":0.26089999079704285},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2590000033378601},{"id":"https://openalex.org/C60782215","wikidata":"https://www.wikidata.org/wiki/Q3333679","display_name":"Probabilistic method","level":3,"score":0.2500999867916107}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.06915","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.06915","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.06915","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.06915","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Modern":[0],"AI":[1],"systems":[2],"are":[3,127],"being":[4],"deployed":[5],"in":[6,78,112],"complex":[7],"domains":[8],"such":[9],"as":[10,36,48],"medicine,":[11],"science,":[12],"and":[13,29,52],"law,":[14],"where":[15],"it":[16],"is":[17],"important":[18],"that":[19],"they":[20],"not":[21],"only":[22],"produce":[23,91],"correct":[24],"answers,":[25],"but":[26],"also":[27],"represent":[28],"update":[30,66],"uncertain":[31],"beliefs":[32,69],"about":[33],"the":[34,42,54,60,103,120,125],"world":[35,126],"new":[37],"evidence":[38,83],"arrives.":[39],"We":[40],"introduce":[41],"novel":[43],"technique":[44],"of":[45,63,88,114,124],"studying":[46],"LLMs":[47,65,80],"information":[49,55],"processing":[50,56],"rules":[51],"utilize":[53],"gap":[57],"to":[58,97,138],"study":[59],"internal":[61],"(in)consistencies":[62],"how":[64,132],"their":[67,85],"probabilistic":[68,122],"from":[70],"evidence.":[71],"Our":[72],"extensive":[73],"experiments":[74],"evaluate":[75],"multiple":[76],"approaches":[77,90],"which":[79],"can":[81,135],"incorporate":[82],"into":[84],"beliefs.":[86],"Some":[87],"these":[89],"(nearly)":[92],"Bayesian":[93,110],"updates;":[94],"others":[95],"seem":[96],"use":[98],"a":[99],"learned":[100],"heuristic.":[101],"Surprisingly,":[102],"non-Bayesian":[104],"heuristic":[105],"updates":[106],"often":[107],"outperform":[108],"exact":[109],"computation":[111],"terms":[113],"downstream":[115],"task":[116],"performance":[117],"--":[118],"indicating":[119],"LLMs'":[121],"models":[123],"misspecified.":[128],"Lastly,":[129],"we":[130],"show":[131],"our":[133],"measure":[134],"provide":[136],"diagnostics":[137],"identify":[139],"issues":[140],"with":[141],"LLM-powered":[142],"inferential":[143],"systems.":[144]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-05-12T00:00:00"}