{"id":"https://openalex.org/W7208719154","doi":"https://doi.org/10.48550/arxiv.2609.03589","title":"Correlated initialization of deep residual networks","display_name":"Correlated initialization of deep residual networks","publication_year":2026,"publication_date":"2026-09-03","ids":{"openalex":"https://openalex.org/W7208719154","doi":"https://doi.org/10.48550/arxiv.2609.03589"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.03589","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.03589","pdf_url":"https://arxiv.org/pdf/2609.03589","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-sa","license_id":"https://openalex.org/licenses/cc-by-sa","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/2609.03589","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5083763539","display_name":"Felix Benning","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Benning, Felix","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5075482281","display_name":"Ivan Nourdin","orcid":"https://orcid.org/0000-0002-8742-0723","observed_orcids":["https://orcid.org/0000-0002-8742-0723"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nourdin, Ivan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5067117041","display_name":"Giovanni Peccati","orcid":"https://orcid.org/0000-0003-2631-2072","observed_orcids":["https://orcid.org/0000-0003-2631-2072"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Peccati, Giovanni","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":1,"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":0.40230000019073486,"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":0.40230000019073486,"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":0.218299999833107,"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/T10067","display_name":"Stochastic processes and financial applications","score":0.041999999433755875,"subfield":{"id":"https://openalex.org/subfields/2003","display_name":"Finance"},"field":{"id":"https://openalex.org/fields/20","display_name":"Economics, Econometrics and Finance"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/resnet","display_name":"ResNet","score":0.9179999828338623},{"id":"https://openalex.org/keywords/hermite-processes","display_name":"Hermite processes","score":0.921999990940094},{"id":"https://openalex.org/keywords/fractional-brownian-motion","display_name":"fractional Brownian motion","score":0.9490000009536743},{"id":"https://openalex.org/keywords/stationary-gaussian-sequences","display_name":"stationary Gaussian sequences","score":0.6100000143051147},{"id":"https://openalex.org/keywords/critical-scaling","display_name":"critical scaling","score":0.6460000276565552},{"id":"https://openalex.org/keywords/neural-network-initialization","display_name":"neural network initialization","score":0.546999990940094}],"concepts":[{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.8371999859809875},{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.6007999777793884},{"id":"https://openalex.org/C201362023","wikidata":"https://www.wikidata.org/wiki/Q658574","display_name":"Hermite polynomials","level":2,"score":0.5442000031471252},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.5311999917030334},{"id":"https://openalex.org/C51955184","wikidata":"https://www.wikidata.org/wiki/Q1545585","display_name":"Stochastic differential equation","level":2,"score":0.46000000834465027},{"id":"https://openalex.org/C108819105","wikidata":"https://www.wikidata.org/wiki/Q1143293","display_name":"Fractional Brownian motion","level":3,"score":0.44290000200271606},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.4401000142097473},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.43650001287460327},{"id":"https://openalex.org/C78045399","wikidata":"https://www.wikidata.org/wiki/Q11214","display_name":"Differential equation","level":2,"score":0.3946000039577484},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.3686000108718872},{"id":"https://openalex.org/C45473103","wikidata":"https://www.wikidata.org/wiki/Q851503","display_name":"Riccati equation","level":3,"score":0.366100013256073},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.3625999987125397},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.34689998626708984},{"id":"https://openalex.org/C151201525","wikidata":"https://www.wikidata.org/wiki/Q177239","display_name":"Limit (mathematics)","level":2,"score":0.34459999203681946},{"id":"https://openalex.org/C112401455","wikidata":"https://www.wikidata.org/wiki/Q178036","display_name":"Brownian motion","level":2,"score":0.33640000224113464},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.33629998564720154},{"id":"https://openalex.org/C24084028","wikidata":"https://www.wikidata.org/wiki/Q7429781","display_name":"Scaling limit","level":3,"score":0.3240000009536743},{"id":"https://openalex.org/C136886441","wikidata":"https://www.wikidata.org/wiki/Q926129","display_name":"Normalization (sociology)","level":2,"score":0.3082999885082245},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.29789999127388},{"id":"https://openalex.org/C121864883","wikidata":"https://www.wikidata.org/wiki/Q677916","display_name":"Statistical physics","level":1,"score":0.2840999960899353},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.27950000762939453},{"id":"https://openalex.org/C116779161","wikidata":"https://www.wikidata.org/wiki/Q7307228","display_name":"Reflected Brownian motion","level":5,"score":0.25690001249313354}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.03589","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.03589","pdf_url":"https://arxiv.org/pdf/2609.03589","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-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.03589","doi":"https://doi.org/10.48550/arxiv.2609.03589","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.03589","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:2609.03589","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.03589","pdf_url":"https://arxiv.org/pdf/2609.03589","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-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G6914970730","display_name":null,"funder_award_id":"O22/17372844/FraMStA","funder_id":"https://openalex.org/F4320321038","funder_display_name":"Fonds National de la Recherche Luxembourg"}],"funders":[{"id":"https://openalex.org/F4320310700","display_name":"Universit\u00e9 du Luxembourg","ror":"https://ror.org/036x5ad56"},{"id":"https://openalex.org/F4320321038","display_name":"Fonds National de la Recherche Luxembourg","ror":"https://ror.org/039z13y21"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7208719154.pdf","grobid_xml":"https://content.openalex.org/works/W7208719154.grobid-xml"},"referenced_works_count":17,"referenced_works":["https://openalex.org/W48692188","https://openalex.org/W617501864","https://openalex.org/W1972736027","https://openalex.org/W1979283834","https://openalex.org/W2005714749","https://openalex.org/W2079974485","https://openalex.org/W2600297185","https://openalex.org/W2619945484","https://openalex.org/W2795427234","https://openalex.org/W2802739963","https://openalex.org/W3144496105","https://openalex.org/W4232171345","https://openalex.org/W4240508093","https://openalex.org/W4249571512","https://openalex.org/W4289345622","https://openalex.org/W4319586203","https://openalex.org/W4383113364"],"related_works":["https://openalex.org/W4282971804","https://openalex.org/W3040629246","https://openalex.org/W7217846722","https://openalex.org/W2966173098","https://openalex.org/W4414595198","https://openalex.org/W2980034437","https://openalex.org/W2755231872","https://openalex.org/W3165175542","https://openalex.org/W7223293282","https://openalex.org/W3100971012"],"abstract_inverted_index":{"We":[0,133],"study":[1],"the":[2,37,47,57,62,88,92,108,113,117,125,128,136,146,152,156,160,167,173,182,193],"large-depth":[3],"behavior":[4],"of":[5,23,64,94,148,155,166,192,195,203],"residual":[6],"networks":[7],"whose":[8],"weights":[9],"are":[10,142],"correlated":[11,31,54],"across":[12],"layers":[13],"at":[14],"initialization.":[15,55],"Our":[16,197],"results":[17,205],"confirm":[18],"and":[19,46,139,163],"extend":[20],"a":[21,65,69,82,95,101,201,207],"conjecture":[22],"Marion":[24],"et":[25],"al.":[26],"[2025],":[27],"according":[28],"to":[29,68,107,189],"which":[30,123],"initializations":[32],"should":[33],"interpolate":[34],"continuously":[35],"between":[36],"Brownian":[38,110,187],"stochastic":[39],"differential":[40,49,97,213],"equation":[41,50,98],"arising":[42,51],"from":[43,52,61],"independent":[44],"initialization":[45,58,118,168],"ordinary":[48],"perfectly":[53],"When":[56],"is":[59,91,124,185],"obtained":[60],"application":[63],"feature":[66,114,157],"function":[67,115],"stationary":[70],"Gaussian":[71],"sequence":[72],"with":[73,151],"regularly":[74],"varying":[75],"correlation,":[76],"we":[77],"prove":[78],"that":[79,87,135],"there":[80],"exists":[81],"unique":[83],"critical":[84,137],"scaling":[85,138],"such":[86],"infinite-depth":[89],"limit":[90,141],"solution":[93],"Young":[96,212],"driven":[99],"by":[100,145],"Hermite":[102,104,120,153,164],"process.":[103],"processes":[105],"reduce":[106],"fractional":[109],"motion":[111],"if":[112],"generating":[116],"has":[119],"rank":[121,154,165],"one,":[122],"case":[126],"for":[127,131,211],"identity":[129],"function,":[130],"example.":[132],"show":[134],"asymptotic":[140,174,183],"uniquely":[143],"determined":[144],"decay":[147],"correlations":[149],"together":[150],"function.":[158],"Consequently,":[159],"correlation":[161],"structure":[162],"represent":[169],"meaningful":[170],"hyperparameters":[171],"in":[172,215],"regime.":[175],"By":[176],"contrast,":[177],"under":[178],"finite-variance":[179],"iid":[180],"initialization,":[181],"driver":[184],"universally":[186],"up":[188],"normalization":[190],"regardless":[191],"choice":[194],"distribution.":[196],"proofs":[198],"rely":[199],"on":[200],"collection":[202],"novel":[204],"establishing":[206],"robust":[208],"stability":[209],"theory":[210],"equations":[214],"Banach":[216],"spaces.":[217]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-05T00:00:00"}