{"id":"https://openalex.org/W7203674978","doi":"https://doi.org/10.48550/arxiv.2608.14664","title":"Quantifying Depth Sufficiency in Residual Neural Networks: A First-Order Criterion","display_name":"Quantifying Depth Sufficiency in Residual Neural Networks: A First-Order Criterion","publication_year":2026,"publication_date":"2026-08-01","ids":{"openalex":"https://openalex.org/W7203674978","doi":"https://doi.org/10.48550/arxiv.2608.14664"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.14664","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.14664","pdf_url":"https://arxiv.org/pdf/2608.14664","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":"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/2608.14664","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5143904182","display_name":"Zeyu Liu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I19820366","display_name":"Chinese Academy of Sciences","ror":"https://ror.org/034t30j35","country_code":"CN","type":"government","lineage":["https://openalex.org/I19820366"]},{"id":"https://openalex.org/I4210090176","display_name":"Institute of Computing Technology","ror":"https://ror.org/0090r4d87","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Liu, Zeyu","raw_affiliation_strings":["Institute of Computing Technology , Chinese Academy of Sciences , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Computing Technology , Chinese Academy of Sciences , Beijing , China","institution_ids":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073912741","display_name":"Jinhao Zhang","orcid":"https://orcid.org/0000-0002-3661-1845","observed_orcids":["https://orcid.org/0000-0002-3661-1845"]},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Jinhao","raw_affiliation_strings":["Beijing University of Posts and Telecommunications , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications , Beijing , China","institution_ids":["https://openalex.org/I139759216"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5122939352","display_name":"Yunquan Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I19820366","display_name":"Chinese Academy of Sciences","ror":"https://ror.org/034t30j35","country_code":"CN","type":"government","lineage":["https://openalex.org/I19820366"]},{"id":"https://openalex.org/I4210090176","display_name":"Institute of Computing Technology","ror":"https://ror.org/0090r4d87","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Yunquan","raw_affiliation_strings":["Institute of Computing Technology , Chinese Academy of Sciences , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Computing Technology , Chinese Academy of Sciences , Beijing , China","institution_ids":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5143900040","display_name":"Guangming Tan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I19820366","display_name":"Chinese Academy of Sciences","ror":"https://ror.org/034t30j35","country_code":"CN","type":"government","lineage":["https://openalex.org/I19820366"]},{"id":"https://openalex.org/I4210090176","display_name":"Institute of Computing Technology","ror":"https://ror.org/0090r4d87","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Tan, Guangming","raw_affiliation_strings":["Institute of Computing Technology , Chinese Academy of Sciences , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Computing Technology , Chinese Academy of Sciences , Beijing , China","institution_ids":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5147391485","display_name":"Xiang Gao","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210123185","display_name":"Zhejiang Lab","ror":"https://ror.org/02m2h7991","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210123185"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Gao, Xiang","raw_affiliation_strings":["High Performance Computing Facility Research Center , Zhejiang Lab , Hangzhou , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"High Performance Computing Facility Research Center , Zhejiang Lab , Hangzhou , China","institution_ids":["https://openalex.org/I4210123185"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048398271","display_name":"Fangming Liu","orcid":"https://orcid.org/0000-0002-8570-1345","observed_orcids":["https://orcid.org/0000-0002-8570-1345"]},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Liu, Fangming","raw_affiliation_strings":["Huazhong University of Science and Technology , Wuhan , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology , Wuhan , China","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5010944123","display_name":"Daning Cheng","orcid":"https://orcid.org/0000-0002-9217-6980","observed_orcids":["https://orcid.org/0000-0002-9217-6980"]},"institutions":[{"id":"https://openalex.org/I19820366","display_name":"Chinese Academy of Sciences","ror":"https://ror.org/034t30j35","country_code":"CN","type":"government","lineage":["https://openalex.org/I19820366"]},{"id":"https://openalex.org/I4210090176","display_name":"Institute of Computing Technology","ror":"https://ror.org/0090r4d87","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Cheng, Daning","raw_affiliation_strings":["Institute of Computing Technology , Chinese Academy of Sciences , Beijing , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute of Computing Technology , Chinese Academy of Sciences , Beijing , China","institution_ids":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"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/T10320","display_name":"Neural Networks and Applications","score":0.49950000643730164,"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/T10320","display_name":"Neural Networks and Applications","score":0.49950000643730164,"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":0.2305999994277954,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.12300000339746475,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/resnet","display_name":"ResNet","score":0.9100000262260437},{"id":"https://openalex.org/keywords/first-order-methods","display_name":"first-order methods","score":0.5479999780654907},{"id":"https://openalex.org/keywords/gpt-2","display_name":"GPT-2","score":0.8640000224113464},{"id":"https://openalex.org/keywords/pythia","display_name":"Pythia","score":0.7089999914169312}],"concepts":[{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.9154000282287598},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.6840999722480774},{"id":"https://openalex.org/C138187205","wikidata":"https://www.wikidata.org/wiki/Q131251","display_name":"Tangent","level":2,"score":0.5471000075340271},{"id":"https://openalex.org/C190470478","wikidata":"https://www.wikidata.org/wiki/Q2370229","display_name":"Invariant (physics)","level":2,"score":0.4652999937534332},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.447299987077713},{"id":"https://openalex.org/C2780012671","wikidata":"https://www.wikidata.org/wiki/Q1343870","display_name":"Corollary","level":2,"score":0.4339999854564667},{"id":"https://openalex.org/C9930424","wikidata":"https://www.wikidata.org/wiki/Q7426587","display_name":"Saturation (graph theory)","level":2,"score":0.40560001134872437},{"id":"https://openalex.org/C191795146","wikidata":"https://www.wikidata.org/wiki/Q3878446","display_name":"Norm (philosophy)","level":2,"score":0.384799987077713},{"id":"https://openalex.org/C2776291640","wikidata":"https://www.wikidata.org/wiki/Q2912517","display_name":"Value (mathematics)","level":2,"score":0.3682999908924103},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3549000024795532},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3458000123500824},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.32170000672340393},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.31619998812675476},{"id":"https://openalex.org/C57493831","wikidata":"https://www.wikidata.org/wiki/Q3134666","display_name":"Projection (relational algebra)","level":2,"score":0.3102000057697296},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.2842999994754791},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.274399995803833},{"id":"https://openalex.org/C62354387","wikidata":"https://www.wikidata.org/wiki/Q875399","display_name":"Boundary (topology)","level":2,"score":0.27070000767707825},{"id":"https://openalex.org/C61445026","wikidata":"https://www.wikidata.org/wiki/Q217608","display_name":"Fixed point","level":2,"score":0.2599000036716461}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.14664","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.14664","pdf_url":"https://arxiv.org/pdf/2608.14664","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.14664","doi":"https://doi.org/10.48550/arxiv.2608.14664","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.14664","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:2608.14664","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.14664","pdf_url":"https://arxiv.org/pdf/2608.14664","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7203674978.pdf","grobid_xml":"https://content.openalex.org/works/W7203674978.grobid-xml"},"referenced_works_count":17,"referenced_works":["https://openalex.org/W2178031510","https://openalex.org/W2331143823","https://openalex.org/W2907121943","https://openalex.org/W3001279689","https://openalex.org/W3080717140","https://openalex.org/W3101845875","https://openalex.org/W3133629262","https://openalex.org/W3170925726","https://openalex.org/W4221150198","https://openalex.org/W4303939357","https://openalex.org/W4323112125","https://openalex.org/W4384111505","https://openalex.org/W4393248014","https://openalex.org/W4399151511","https://openalex.org/W4407422992","https://openalex.org/W4412887764","https://openalex.org/W7160955240"],"related_works":["https://openalex.org/W7140305772","https://openalex.org/W3172612647","https://openalex.org/W2964232029","https://openalex.org/W4416372146","https://openalex.org/W4414595198","https://openalex.org/W7223293282","https://openalex.org/W2541674938","https://openalex.org/W4415311869","https://openalex.org/W7169542570","https://openalex.org/W7166693845"],"abstract_inverted_index":{"How":[0],"can":[1],"we":[2],"determine":[3],"whether":[4],"a":[5,17,43,70,122,144],"trained":[6],"neural":[7],"network":[8],"is":[9,55,83],"already":[10],"deep":[11],"enough?":[12],"We":[13,33,51],"study":[14],"this":[15],"under":[16,91],"fixed":[18],"function-preserving":[19],"residual-growth":[20],"protocol":[21],"specifying":[22],"insertion":[23],"locations,":[24],"residual":[25,36,53,78,154],"families,":[26],"zero-output":[27],"initializations,":[28],"and":[29,57,89,112],"zero-state":[30,87],"first-order":[31,35,62,151],"updates.":[32],"define":[34],"depth":[37,60],"saturation":[38],"as":[39,143],"the":[40,116,148],"absence":[41],"of":[42,147,153],"strict":[44],"local":[45,93],"decrease":[46],"from":[47,136],"every":[48],"admissible":[49,77],"insertion.":[50],"prove":[52],"non-degeneracy":[54],"necessary":[56],"sufficient:":[58],"additional":[59],"has":[61],"value":[63,152],"exactly":[64,105],"when":[65],"conditional":[66],"activation":[67],"gradients":[68],"have":[69],"nonzero":[71],"projection":[72],"onto":[73],"at":[74],"least":[75],"one":[76],"tangent":[79,97],"space.":[80,98],"This":[81],"boundary":[82],"shared":[84],"by":[85],"descent-compatible":[86],"updates":[88],"invariant":[90],"regular":[92],"reparameterizations":[94],"preserving":[95],"that":[96],"Under":[99],"residual-signal":[100],"realizability,":[101],"raw":[102],"activation-gradient":[103,118,141],"vanishing":[104],"certifies":[106],"saturation.":[107],"Across":[108],"ResNets,":[109],"GPT-2-style":[110],"models,":[111],"continued-pretrained":[113],"Pythia":[114],"checkpoints,":[115],"maximum":[117],"norm":[119],"decreases":[120],"toward":[121],"low-signal":[123],"regime":[124],"with":[125,134],"depth.":[126,155],"Function-preserving":[127],"growth":[128],"also":[129],"achieves":[130],"converged":[131],"performance":[132],"competitive":[133],"training":[135],"scratch.":[137],"These":[138],"results":[139],"support":[140],"magnitude":[142],"conservative":[145],"diagnostic":[146],"remaining":[149],"empirical":[150]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-08-19T00:00:00"}