{"id":"https://openalex.org/W7220951483","doi":"https://doi.org/10.48550/arxiv.2610.09132","title":"The Impact of Likelihood Tempering on the Limiting Predictive Moments of Variational Bayesian Linear Neural Networks","display_name":"The Impact of Likelihood Tempering on the Limiting Predictive Moments of Variational Bayesian Linear Neural Networks","publication_year":2026,"publication_date":"2026-10-06","ids":{"openalex":"https://openalex.org/W7220951483","doi":"https://doi.org/10.48550/arxiv.2610.09132"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.09132","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.09132","pdf_url":"https://arxiv.org/pdf/2610.09132","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/2610.09132","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5114730060","display_name":"Ian Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Zhang, Ian","raw_affiliation_strings":["Department of Statistical Sciences , University of Toronto","Dunlap Institute for Astronomy and Astrophysics"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Statistical Sciences , University of Toronto","institution_ids":["https://openalex.org/I185261750"]},{"raw_affiliation_string":"Dunlap Institute for Astronomy and Astrophysics","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5070514084","display_name":"Thibault Randrianarisoa","orcid":"https://orcid.org/0000-0002-2681-1093","observed_orcids":["https://orcid.org/0000-0002-2681-1093"]},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Randrianarisoa, Thibault","raw_affiliation_strings":["Department of Statistical Sciences , University of Toronto"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Statistical Sciences , University of Toronto","institution_ids":["https://openalex.org/I185261750"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I185261750"],"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":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12814","display_name":"Gaussian Processes and Bayesian Inference","score":0.9984999895095825,"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/T12814","display_name":"Gaussian Processes and Bayesian Inference","score":0.9984999895095825,"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/T10320","display_name":"Neural Networks and Applications","score":0.0006000000284984708,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":9.999999747378752e-05,"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/variational-bayesian-inference","display_name":"variational inference","score":0.8009999990463257},{"id":"https://openalex.org/keywords/linear-neural-networks","display_name":"linear neural networks","score":0.6909999847412109},{"id":"https://openalex.org/keywords/infinite-width-limit","display_name":"infinite-width limit","score":0.7070000171661377},{"id":"https://openalex.org/keywords/gaussian-processes","display_name":"Gaussian processes","score":0.5210000276565552},{"id":"https://openalex.org/keywords/phase-transitions","display_name":"phase transitions","score":0.593999981880188},{"id":"https://openalex.org/keywords/temperature-schedule","display_name":"temperature schedule","score":0.31700000166893005}],"concepts":[{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.5888000130653381},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.5830000042915344},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.5159000158309937},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.49900001287460327},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.49630001187324524},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.4564000070095062},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.40220001339912415},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.40070000290870667},{"id":"https://openalex.org/C188198153","wikidata":"https://www.wikidata.org/wiki/Q1613840","display_name":"Limiting","level":2,"score":0.3776000142097473},{"id":"https://openalex.org/C57830394","wikidata":"https://www.wikidata.org/wiki/Q278079","display_name":"Posterior probability","level":3,"score":0.3621000051498413},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.35899999737739563},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3538999855518341},{"id":"https://openalex.org/C65778772","wikidata":"https://www.wikidata.org/wiki/Q12345341","display_name":"Asymptotic distribution","level":3,"score":0.35030001401901245},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.35010001063346863},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.34709998965263367},{"id":"https://openalex.org/C177769412","wikidata":"https://www.wikidata.org/wiki/Q278090","display_name":"Prior probability","level":3,"score":0.34619998931884766},{"id":"https://openalex.org/C2777727622","wikidata":"https://www.wikidata.org/wiki/Q5251772","display_name":"Degeneracy (biology)","level":2,"score":0.3425999879837036},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.3280999958515167},{"id":"https://openalex.org/C163175372","wikidata":"https://www.wikidata.org/wiki/Q3339222","display_name":"Linear model","level":2,"score":0.3093999922275543},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.3010999858379364},{"id":"https://openalex.org/C184050105","wikidata":"https://www.wikidata.org/wiki/Q273163","display_name":"Isotropy","level":2,"score":0.29840001463890076},{"id":"https://openalex.org/C161999928","wikidata":"https://www.wikidata.org/wiki/Q4556320","display_name":"Variational inequality","level":2,"score":0.2874000072479248},{"id":"https://openalex.org/C43555835","wikidata":"https://www.wikidata.org/wiki/Q2300258","display_name":"Conditional probability distribution","level":2,"score":0.2872999906539917},{"id":"https://openalex.org/C37903108","wikidata":"https://www.wikidata.org/wiki/Q4874474","display_name":"Bayesian linear regression","level":4,"score":0.28679999709129333},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.2786000072956085},{"id":"https://openalex.org/C205606062","wikidata":"https://www.wikidata.org/wiki/Q5249645","display_name":"Decoupling (probability)","level":2,"score":0.27469998598098755},{"id":"https://openalex.org/C110121322","wikidata":"https://www.wikidata.org/wiki/Q865811","display_name":"Distribution (mathematics)","level":2,"score":0.27320000529289246},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.2646999955177307},{"id":"https://openalex.org/C2778136018","wikidata":"https://www.wikidata.org/wiki/Q10350689","display_name":"Predictive power","level":2,"score":0.26190000772476196},{"id":"https://openalex.org/C2777027219","wikidata":"https://www.wikidata.org/wiki/Q1284190","display_name":"Constant (computer programming)","level":2,"score":0.2612999975681305},{"id":"https://openalex.org/C207467116","wikidata":"https://www.wikidata.org/wiki/Q4385666","display_name":"Inverse","level":2,"score":0.25589999556541443},{"id":"https://openalex.org/C121864883","wikidata":"https://www.wikidata.org/wiki/Q677916","display_name":"Statistical physics","level":1,"score":0.25110000371932983}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.09132","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.09132","pdf_url":"https://arxiv.org/pdf/2610.09132","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":"doi:10.48550/arxiv.2610.09132","doi":"https://doi.org/10.48550/arxiv.2610.09132","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.09132","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":"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2610.09132","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.09132","pdf_url":"https://arxiv.org/pdf/2610.09132","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/W7220951483.pdf","grobid_xml":"https://content.openalex.org/works/W7220951483.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4383722908","https://openalex.org/W3172357953","https://openalex.org/W2784590060","https://openalex.org/W4221167850","https://openalex.org/W2281271687","https://openalex.org/W3037219792","https://openalex.org/W2966381196","https://openalex.org/W7131873354","https://openalex.org/W4285044940","https://openalex.org/W3181595108"],"abstract_inverted_index":{"In":[0],"wide":[1],"Bayesian":[2],"neural":[3,129],"networks,":[4],"Gaussian":[5,97,131],"mean-field":[6],"variational":[7,28],"inference":[8],"is":[9,57,145],"prone":[10],"to":[11,32,47,59,78],"\"prior":[12],"dominance\":":[13],"the":[14,20,23,27,33,37,42,48,61,88,101,108,127,135,139,147,158,178,186,198,217],"Kullback-Leibler":[15],"(KL)":[16],"regularization":[17],"term":[18,63],"of":[19,107,138,211],"ELBO":[21],"outweighs":[22],"expected":[24],"log-likelihood,":[25],"and":[26,82,123,150,176,202],"predictive":[29,35,103,148],"distribution":[30,104],"collapses":[31],"prior":[34,163,191],"as":[36,119],"width":[38],"$M$":[39,77],"grows.":[40],"Tempering":[41],"likelihood,":[43],"by":[44,64],"raising":[45],"it":[46,125],"power":[49],"$1/T$":[50],"for":[51,181,193,200,204],"a":[52,84],"temperature":[53],"$T":[54,110],"<":[55,195],"1$,":[56,196],"equivalent":[58],"scaling":[60],"KL":[62],"$T$.":[65],"We":[66],"ask":[67],"in":[68],"this":[69,80],"paper":[70],"how":[71],"fast":[72],"$T$":[73],"must":[74],"decrease":[75],"with":[76,95,113,126],"counteract":[79],"degeneracy":[81],"strike":[83],"good":[85],"balance":[86],"between":[87],"two":[89],"terms.":[90],"For":[91],"single-hidden-layer":[92],"linear":[93],"networks":[94],"isotropic":[96],"priors,":[98],"we":[99],"derive":[100],"limiting":[102,159,187],"under":[105],"schedules":[106],"form":[109],"=":[111,167],"\u03c4/M^{c}$,":[112],"constants":[114],"$\u03c4,":[115],"c":[116],">":[117,183,206],"0$,":[118],"$M":[120],"\\to":[121],"\\infty$":[122],"compare":[124],"untempered":[128],"network":[130],"process":[132],"(NNGP)":[133],"posterior,":[134],"infinite-width":[136],"limit":[137],"exact":[140],"posterior.":[141],"Our":[142],"main":[143],"result":[144],"that":[146],"expectation":[149,160,220],"variance":[151,188],"undergo":[152],"phase":[153],"transitions":[154],"at":[155,165],"different":[156],"scales:":[157],"leaves":[161],"its":[162,190,222],"value":[164,192],"$c":[166,182,194,205],"1/2$,":[168,184],"once":[169],"$\u03c4$":[170],"falls":[171],"below":[172],"an":[173],"explicit":[174],"threshold,":[175],"equals":[177],"least-squares":[179],"prediction":[180],"whereas":[185],"keeps":[189],"matches":[197],"NNGP's":[199],"$c=1$,":[201],"vanishes":[203],"1$.":[207],"With":[208],"suitable":[209],"choices":[210],"$\u03c4,c$,":[212],"one":[213],"can":[214],"recover":[215],"either":[216],"NNGP":[218],"posterior":[219],"or":[221],"variance.":[223]},"counts_by_year":[],"updated_date":"2026-11-06T12:05:46.655746","created_date":"2026-10-09T00:00:00"}