{"id":"https://openalex.org/W7166797992","doi":"https://doi.org/10.48550/arxiv.2606.30850","title":"BayesBench: Evaluating LLM Belief Trajectories Under Multi-Turn Evidence Accumulation","display_name":"BayesBench: Evaluating LLM Belief Trajectories Under Multi-Turn Evidence Accumulation","publication_year":2026,"publication_date":"2026-06-29","ids":{"openalex":"https://openalex.org/W7166797992","doi":"https://doi.org/10.48550/arxiv.2606.30850"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.30850","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.30850","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.30850","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139757884","display_name":"Ankur Samanta","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Samanta, Ankur","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102706189","display_name":"Akshayaa Magesh","orcid":"https://orcid.org/0000-0002-4627-9321"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Magesh, Akshayaa","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037346243","display_name":"Tal Lancewicki","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lancewicki, Tal","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137708499","display_name":"A Jain","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jain, Ayush","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139818161","display_name":"Youliang Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Youliang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139770917","display_name":"Paul Sajda","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sajda, Paul","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5072012496","display_name":"Kaveh Hassani","orcid":"https://orcid.org/0000-0001-9162-9442"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hassani, Kaveh","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030766536","display_name":"Aditya Modi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Modi, Aditya","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139788004","display_name":"Daniel R. Jiang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiang, Daniel R.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5090891199","display_name":"Yonathan Efroni","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Efroni, Yonathan","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":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/T10028","display_name":"Topic Modeling","score":0.5367000102996826,"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":0.5367000102996826,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.0731000006198883,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T12031","display_name":"Speech and dialogue systems","score":0.05689999833703041,"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/bayesian-inference","display_name":"Bayesian inference","score":0.6455000042915344},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6324999928474426},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.6309999823570251},{"id":"https://openalex.org/keywords/latent-variable","display_name":"Latent variable","score":0.5475999712944031},{"id":"https://openalex.org/keywords/outcome","display_name":"Outcome (game theory)","score":0.5034999847412109},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.4609000086784363},{"id":"https://openalex.org/keywords/latent-variable-model","display_name":"Latent variable model","score":0.415800005197525},{"id":"https://openalex.org/keywords/bayesian-statistics","display_name":"Bayesian statistics","score":0.40880000591278076},{"id":"https://openalex.org/keywords/semantic-reasoner","display_name":"Semantic reasoner","score":0.40290001034736633}],"concepts":[{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.6455000042915344},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6324999928474426},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.6309999823570251},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5569999814033508},{"id":"https://openalex.org/C51167844","wikidata":"https://www.wikidata.org/wiki/Q4422623","display_name":"Latent variable","level":2,"score":0.5475999712944031},{"id":"https://openalex.org/C148220186","wikidata":"https://www.wikidata.org/wiki/Q7111912","display_name":"Outcome (game theory)","level":2,"score":0.5034999847412109},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.49900001287460327},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.47749999165534973},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.4609000086784363},{"id":"https://openalex.org/C65965080","wikidata":"https://www.wikidata.org/wiki/Q1806885","display_name":"Latent variable model","level":3,"score":0.415800005197525},{"id":"https://openalex.org/C101112237","wikidata":"https://www.wikidata.org/wiki/Q4874481","display_name":"Bayesian statistics","level":4,"score":0.40880000591278076},{"id":"https://openalex.org/C9616225","wikidata":"https://www.wikidata.org/wiki/Q3929429","display_name":"Semantic reasoner","level":2,"score":0.40290001034736633},{"id":"https://openalex.org/C149782125","wikidata":"https://www.wikidata.org/wiki/Q160039","display_name":"Econometrics","level":1,"score":0.3991999924182892},{"id":"https://openalex.org/C90329073","wikidata":"https://www.wikidata.org/wiki/Q914232","display_name":"Ask price","level":2,"score":0.3384000062942505},{"id":"https://openalex.org/C108650721","wikidata":"https://www.wikidata.org/wiki/Q1783253","display_name":"Counterfactual thinking","level":2,"score":0.31850001215934753},{"id":"https://openalex.org/C207201462","wikidata":"https://www.wikidata.org/wiki/Q182505","display_name":"Bayes' theorem","level":3,"score":0.3154999911785126},{"id":"https://openalex.org/C162376815","wikidata":"https://www.wikidata.org/wiki/Q2158281","display_name":"Frequentist inference","level":4,"score":0.3089999854564667},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.2994999885559082},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.29649999737739563},{"id":"https://openalex.org/C182365436","wikidata":"https://www.wikidata.org/wiki/Q50701","display_name":"Variable (mathematics)","level":2,"score":0.2768000066280365},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.27230000495910645},{"id":"https://openalex.org/C142291917","wikidata":"https://www.wikidata.org/wiki/Q4165283","display_name":"Bayes factor","level":4,"score":0.271699994802475},{"id":"https://openalex.org/C75455068","wikidata":"https://www.wikidata.org/wiki/Q1455566","display_name":"Frequentist probability","level":3,"score":0.26750001311302185},{"id":"https://openalex.org/C917703","wikidata":"https://www.wikidata.org/wiki/Q7239668","display_name":"Predictive inference","level":5,"score":0.26600000262260437},{"id":"https://openalex.org/C201717286","wikidata":"https://www.wikidata.org/wiki/Q938185","display_name":"Rationality","level":2,"score":0.2547999918460846},{"id":"https://openalex.org/C70727504","wikidata":"https://www.wikidata.org/wiki/Q1806878","display_name":"Latent class model","level":2,"score":0.2508000135421753}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.30850","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.30850","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.30850","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.30850","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"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":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"are":[4,129],"typically":[5],"deployed":[6],"in":[7,52,74],"multi-turn":[8,75],"conversations,":[9],"where":[10,96,109,127],"each":[11],"turn":[12],"provides":[13],"new":[14],"evidence":[15,41,154],"that":[16,32,85],"should":[17],"reduce":[18],"epistemic":[19],"uncertainty":[20],"about":[21,38,115,188],"their":[22],"environment.":[23],"Acting":[24],"rationally":[25,185],"then":[26],"requires":[27],"inferring":[28,178],"the":[29,48,97,110,139,143,160,189],"unobserved":[30],"quantities":[31],"govern":[33],"it":[34,183],"and":[35,77,122,142,153,181],"updating":[36],"beliefs":[37,114,187],"them":[39],"as":[40],"accumulates.":[42],"Yet":[43],"most":[44],"evaluations":[45],"only":[46],"score":[47],"model's":[49],"final-turn":[50],"answer":[51],"a":[53,70,80,116,132,175],"single-turn":[54],"format,":[55],"leaving":[56],"this":[57,87],"process":[58],"unexamined.":[59],"We":[60],"ask":[61],"how":[62],"closely":[63],"LLMs'":[64],"belief":[65],"updates":[66,157],"match":[67],"those":[68],"of":[69,82],"rational":[71],"Bayesian":[72,94,107,125,161],"reasoner":[73],"settings,":[76],"introduce":[78],"BayesBench,":[79],"suite":[81],"simulation":[83],"environments":[84],"probe":[86],"across":[88],"three":[89],"progressively":[90],"complex":[91],"tasks:":[92],"(i)":[93],"estimation,":[95],"model":[98,111],"infers":[99],"an":[100],"unknown":[101],"parameter":[102],"from":[103],"sequential":[104],"evidence;":[105],"(ii)":[106],"prediction,":[108,126,173],"turns":[112],"inferred":[113],"latent":[117,140,151,179],"variable":[118],"into":[119],"outcome":[120],"forecasts;":[121],"(iii)":[123],"latent-framed":[124],"observations":[128],"filtered":[130],"through":[131],"user-persona":[133],"framing,":[134],"requiring":[135],"joint":[136],"inference":[137,152],"over":[138,170],"state":[141],"persona.":[144],"Across":[145],"seven":[146],"LLMs":[147],"(3B--70B),":[148],"scaling":[149],"improves":[150],"accumulation,":[155],"with":[156],"occasionally":[158],"matching":[159],"posterior.":[162],"However,":[163],"these":[164],"gains":[165],"do":[166],"not":[167],"reliably":[168],"carry":[169],"to":[171,184],"downstream":[172],"exposing":[174],"gap":[176],"between":[177],"structure":[180],"using":[182],"update":[186],"target":[190],"outcome.":[191]},"counts_by_year":[],"updated_date":"2026-07-02T06:18:51.028212","created_date":"2026-07-02T00:00:00"}