{"id":"https://openalex.org/W7162785606","doi":"https://doi.org/10.48550/arxiv.2605.29684","title":"Kernel Renormalization in Bayesian Deep Neural Networks: the Equivalent Wishart Ansatz in the Proportional Regime","display_name":"Kernel Renormalization in Bayesian Deep Neural Networks: the Equivalent Wishart Ansatz in the Proportional Regime","publication_year":2026,"publication_date":"2026-05-28","ids":{"openalex":"https://openalex.org/W7162785606","doi":"https://doi.org/10.48550/arxiv.2605.29684"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.29684","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29684","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,"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.2605.29684","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5055620681","display_name":"Paolo Baglioni","orcid":"https://orcid.org/0000-0002-9214-6669"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Baglioni, Paolo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021471102","display_name":"Christian Keup","orcid":"https://orcid.org/0000-0001-6258-8135"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Keup, Christian","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137371136","display_name":"Vincenzo Zimbardo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zimbardo, Vincenzo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137328580","display_name":"Rosalba Pacelli","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pacelli, Rosalba","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066787935","display_name":"A. Vezzani","orcid":"https://orcid.org/0000-0002-9779-6150"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Vezzani, Alessandro","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5038411636","display_name":"Raffaella Burioni","orcid":"https://orcid.org/0000-0003-4648-6838"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Burioni, Raffaella","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5021125060","display_name":"Pietro Rotondo","orcid":"https://orcid.org/0000-0001-6745-6166"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rotondo, Pietro","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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.4756999909877777,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.4756999909877777,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.17669999599456787,"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/T12676","display_name":"Machine Learning and ELM","score":0.06610000133514404,"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/perceptron","display_name":"Perceptron","score":0.5698999762535095},{"id":"https://openalex.org/keywords/kernel","display_name":"Kernel (algebra)","score":0.5206000208854675},{"id":"https://openalex.org/keywords/ansatz","display_name":"Ansatz","score":0.5105999708175659},{"id":"https://openalex.org/keywords/bayesian-probability","display_name":"Bayesian probability","score":0.4505000114440918},{"id":"https://openalex.org/keywords/limit","display_name":"Limit (mathematics)","score":0.40689998865127563},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.3921999931335449},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.3840000033378601},{"id":"https://openalex.org/keywords/renormalization","display_name":"Renormalization","score":0.383899986743927},{"id":"https://openalex.org/keywords/hyperparameter","display_name":"Hyperparameter","score":0.3815000057220459}],"concepts":[{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.6317999958992004},{"id":"https://openalex.org/C60908668","wikidata":"https://www.wikidata.org/wiki/Q690207","display_name":"Perceptron","level":3,"score":0.5698999762535095},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.5206000208854675},{"id":"https://openalex.org/C130979935","wikidata":"https://www.wikidata.org/wiki/Q568954","display_name":"Ansatz","level":2,"score":0.5105999708175659},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.4505000114440918},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.41519999504089355},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.40689998865127563},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3921999931335449},{"id":"https://openalex.org/C121864883","wikidata":"https://www.wikidata.org/wiki/Q677916","display_name":"Statistical physics","level":1,"score":0.38510000705718994},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3840000033378601},{"id":"https://openalex.org/C166124518","wikidata":"https://www.wikidata.org/wiki/Q1047702","display_name":"Renormalization","level":2,"score":0.383899986743927},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.3815000057220459},{"id":"https://openalex.org/C188441871","wikidata":"https://www.wikidata.org/wiki/Q7554146","display_name":"Softmax function","level":3,"score":0.37619999051094055},{"id":"https://openalex.org/C122280245","wikidata":"https://www.wikidata.org/wiki/Q620622","display_name":"Kernel method","level":3,"score":0.33649998903274536},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.3303999900817871},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.32089999318122864},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.31360000371932983},{"id":"https://openalex.org/C57691317","wikidata":"https://www.wikidata.org/wiki/Q1289248","display_name":"Scalar (mathematics)","level":2,"score":0.3086000084877014},{"id":"https://openalex.org/C207201462","wikidata":"https://www.wikidata.org/wiki/Q182505","display_name":"Bayes' theorem","level":3,"score":0.30559998750686646},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2962999939918518},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.29339998960494995},{"id":"https://openalex.org/C68532491","wikidata":"https://www.wikidata.org/wiki/Q1203669","display_name":"Renormalization group","level":2,"score":0.2766000032424927},{"id":"https://openalex.org/C24084028","wikidata":"https://www.wikidata.org/wiki/Q7429781","display_name":"Scaling limit","level":3,"score":0.2752000093460083},{"id":"https://openalex.org/C97937538","wikidata":"https://www.wikidata.org/wiki/Q199691","display_name":"Laplace transform","level":2,"score":0.2639000117778778},{"id":"https://openalex.org/C2778401447","wikidata":"https://www.wikidata.org/wiki/Q7140637","display_name":"Partition function (quantum field theory)","level":2,"score":0.25859999656677246},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.2572000026702881},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2565000057220459},{"id":"https://openalex.org/C169903167","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Test set","level":2,"score":0.25440001487731934},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.2540999948978424}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.29684","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29684","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,"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.2605.29684","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.29684","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,"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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"The":[0],"scaling":[1],"limit":[2,135],"where":[3],"both":[4],"the":[5,8,13,23,26,63,87,92,108,114,133,148,171,189],"size":[6],"of":[7,16,66,71,91,96,111,120,170,192,219],"training":[9],"set":[10],"$P$":[11],"and":[12,202],"width":[14],"$N$":[15],"a":[17,103,121,156],"deep":[18,47,194],"neural":[19,195],"network":[20],"grow":[21],"at":[22,139],"same":[24],"rate,":[25],"so-called":[27],"proportional-width":[28],"regime,":[29],"has":[30,50],"been":[31],"intensely":[32],"studied":[33],"for":[34,107],"shallow,":[35],"single-hidden-layer":[36],"networks.":[37],"However,":[38],"extending":[39],"these":[40],"non-perturbative":[41],"results":[42],"from":[43,188],"shallow":[44],"architectures":[45,152],"to":[46,61,85,101,150,164,177],"non-linear":[48],"networks":[49,196],"proven":[51],"very":[52,211],"challenging.":[53],"Here":[54],"we":[55,154],"present":[56],"an":[57,81],"effective":[58,183],"approximate":[59],"approach":[60,149],"predict":[62],"generalization":[64],"performance":[65],"Bayesian":[67,190],"multi-layer":[68],"perceptrons":[69],"(MLPs)":[70],"fixed":[72],"depth":[73],"$L$":[74,141],"on":[75,205],"arbitrary":[76],"high-dimensional":[77],"data.":[78],"We":[79,180],"propose":[80],"equivalent":[82],"Wishart":[83],"Ansatz":[84],"capture":[86],"dominant":[88],"stochastic":[89],"fluctuations":[90],"hierarchical":[93,157],"empirical":[94],"kernels":[95],"MLPs.":[97],"This":[98],"allows":[99,163],"us":[100],"perform":[102],"large":[104],"deviation":[105],"analysis":[106],"partition":[109],"function":[110],"MLPs":[112],"in":[113,118,132,138,174],"proportional":[115,134],"limit,":[116],"expressed":[117],"terms":[119],"renormalized":[122],"NNGP":[123],"kernel.":[124],"In":[125],"this":[126],"description,":[127],"even":[128],"strong":[129],"representation":[130],"learning":[131],"is":[136],"encoded":[137],"most":[140],"scalar":[142],"order":[143],"parameters,":[144],"determined":[145],"self-consistently.":[146],"Extending":[147],"convolutional":[151],"(CNNs),":[153],"identify":[155],"local":[158],"kernel":[159,173],"renormalization":[160],"mechanism,":[161],"which":[162],"quantify":[165],"more":[166],"complex":[167],"data-dependent":[168],"transformations":[169],"large-width":[172],"CNNs":[175],"due":[176],"finite-width":[178],"effects.":[179],"test":[181],"our":[182],"theory":[184],"against":[185],"sampling":[186],"experiments":[187],"posterior":[191],"finite":[193],"with":[197,215],"depths":[198],"$L":[199],"\\sim":[200],"O(10)$":[201],"$P\\sim":[203],"O(10^3)$":[204],"classic":[206],"benchmark":[207],"datasets,":[208],"finding":[209],"overall":[210],"good":[212],"agreement":[213],"together":[214],"two":[216],"distinct":[217],"types":[218],"systematic":[220],"deviations.":[221]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-05-30T00:00:00"}