{"id":"https://openalex.org/W4393241094","doi":"https://doi.org/10.1109/tpami.2024.3382294","title":"Towards Understanding Convergence and Generalization of AdamW","display_name":"Towards Understanding Convergence and Generalization of AdamW","publication_year":2024,"publication_date":"2024-03-27","ids":{"openalex":"https://openalex.org/W4393241094","doi":"https://doi.org/10.1109/tpami.2024.3382294","pmid":"https://pubmed.ncbi.nlm.nih.gov/38536692"},"language":"en","primary_location":{"id":"doi:10.1109/tpami.2024.3382294","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpami.2024.3382294","pdf_url":null,"source":{"id":"https://openalex.org/S199944782","display_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","issn_l":"0162-8828","issn":["0162-8828","1939-3539","2160-9292"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320439","host_organization_name":"IEEE Computer Society","host_organization_lineage":["https://openalex.org/P4310320439","https://openalex.org/P4310319808"],"host_organization_lineage_names":["IEEE Computer Society","Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","pubmed"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://ink.library.smu.edu.sg/sis_research/8986","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5010883708","display_name":"Pan Zhou","orcid":"https://orcid.org/0000-0003-3400-8943"},"institutions":[{"id":"https://openalex.org/I79891267","display_name":"Singapore Management University","ror":"https://ror.org/050qmg959","country_code":"SG","type":"education","lineage":["https://openalex.org/I79891267"]}],"countries":["SG"],"is_corresponding":true,"raw_author_name":"Pan Zhou","raw_affiliation_strings":["School of Computing and Information Systems, Singapore Management University, Singapore"],"affiliations":[{"raw_affiliation_string":"School of Computing and Information Systems, Singapore Management University, Singapore","institution_ids":["https://openalex.org/I79891267"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048069849","display_name":"Xingyu Xie","orcid":"https://orcid.org/0000-0002-8756-5981"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xingyu Xie","raw_affiliation_strings":["National Key Lab of General AI, School of Intelligence Science and Technology, Peking University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"National Key Lab of General AI, School of Intelligence Science and Technology, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5016399094","display_name":"Zhouchen Lin","orcid":"https://orcid.org/0000-0003-1493-7569"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]},{"id":"https://openalex.org/I4210100255","display_name":"Beijing Academy of Artificial Intelligence","ror":"https://ror.org/016a74861","country_code":"CN","type":"other","lineage":["https://openalex.org/I4210100255"]},{"id":"https://openalex.org/I4210136793","display_name":"Peng Cheng Laboratory","ror":"https://ror.org/03qdqbt06","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210136793"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhouchen Lin","raw_affiliation_strings":["National Key Lab of General AI, School of Intelligence Science and Technology, Peking University, Beijing, China","Institute for Artificial Intelligence, Peking University, China","Peng Cheng Laboratory, China"],"affiliations":[{"raw_affiliation_string":"National Key Lab of General AI, School of Intelligence Science and Technology, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"Institute for Artificial Intelligence, Peking University, China","institution_ids":["https://openalex.org/I4210100255","https://openalex.org/I20231570"]},{"raw_affiliation_string":"Peng Cheng Laboratory, China","institution_ids":["https://openalex.org/I4210136793"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100381753","display_name":"Shuicheng Yan","orcid":"https://orcid.org/0000-0001-8906-3777"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shuicheng Yan","raw_affiliation_strings":["Skywork AI, Mingyang International Center, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Skywork AI, Mingyang International Center, Beijing, China","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5010883708"],"corresponding_institution_ids":["https://openalex.org/I79891267"],"apc_list":null,"apc_paid":null,"fwci":61.1538,"has_fulltext":false,"cited_by_count":180,"citation_normalized_percentile":{"value":0.99983559,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":"46","issue":"9","first_page":"6486","last_page":"6493"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12176","display_name":"Optimization and Packing Problems","score":0.9721999764442444,"subfield":{"id":"https://openalex.org/subfields/2209","display_name":"Industrial and Manufacturing Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12176","display_name":"Optimization and Packing Problems","score":0.9721999764442444,"subfield":{"id":"https://openalex.org/subfields/2209","display_name":"Industrial and Manufacturing Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.960099995136261,"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/T10057","display_name":"Face and Expression Recognition","score":0.9563999772071838,"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/computer-science","display_name":"Computer science","score":0.5749574899673462},{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.5717061161994934},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.5548977851867676},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.5169731974601746},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.34102165699005127},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.2778371572494507}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5749574899673462},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.5717061161994934},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5548977851867676},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.5169731974601746},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.34102165699005127},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2778371572494507},{"id":"https://openalex.org/C50522688","wikidata":"https://www.wikidata.org/wiki/Q189833","display_name":"Economic growth","level":1,"score":0.0},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.1109/tpami.2024.3382294","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpami.2024.3382294","pdf_url":null,"source":{"id":"https://openalex.org/S199944782","display_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","issn_l":"0162-8828","issn":["0162-8828","1939-3539","2160-9292"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310320439","host_organization_name":"IEEE Computer Society","host_organization_lineage":["https://openalex.org/P4310320439","https://openalex.org/P4310319808"],"host_organization_lineage_names":["IEEE Computer Society","Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Pattern Analysis and Machine Intelligence","raw_type":"journal-article"},{"id":"pmid:38536692","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/38536692","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE transactions on pattern analysis and machine intelligence","raw_type":null},{"id":"pmh:oai:ink.library.smu.edu.sg:sis_research-9989","is_oa":true,"landing_page_url":"https://ink.library.smu.edu.sg/sis_research/8986","pdf_url":null,"source":{"id":"https://openalex.org/S4306401925","display_name":"Singapore Management University Institutional Knowledge (InK) (Singapore Management University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I79891267","host_organization_name":"Singapore Management University","host_organization_lineage":["https://openalex.org/I79891267"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"https://doi.org/10.1109/TPAMI.2024.3382294","raw_type":"Journal Article"}],"best_oa_location":{"id":"pmh:oai:ink.library.smu.edu.sg:sis_research-9989","is_oa":true,"landing_page_url":"https://ink.library.smu.edu.sg/sis_research/8986","pdf_url":null,"source":{"id":"https://openalex.org/S4306401925","display_name":"Singapore Management University Institutional Knowledge (InK) (Singapore Management University)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I79891267","host_organization_name":"Singapore Management University","host_organization_lineage":["https://openalex.org/I79891267"],"host_organization_lineage_names":[],"type":"repository"},"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"https://doi.org/10.1109/TPAMI.2024.3382294","raw_type":"Journal Article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","score":0.5799999833106995,"display_name":"Reduced inequalities"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320320751","display_name":"Ministry of Education - Singapore","ror":"https://ror.org/01kcva023"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":68,"referenced_works":["https://openalex.org/W280536263","https://openalex.org/W1530699444","https://openalex.org/W1677182931","https://openalex.org/W1686810756","https://openalex.org/W1994616650","https://openalex.org/W1995562189","https://openalex.org/W2108598243","https://openalex.org/W2194775991","https://openalex.org/W2331050942","https://openalex.org/W2605488176","https://openalex.org/W2626017178","https://openalex.org/W2626325961","https://openalex.org/W2752366553","https://openalex.org/W2768267830","https://openalex.org/W2962915600","https://openalex.org/W2993258424","https://openalex.org/W3094502228","https://openalex.org/W3107436194","https://openalex.org/W3109623941","https://openalex.org/W3116376544","https://openalex.org/W3138582970","https://openalex.org/W3159481202","https://openalex.org/W3166612201","https://openalex.org/W3176776344","https://openalex.org/W4221155355","https://openalex.org/W4244569433","https://openalex.org/W4246193833","https://openalex.org/W4255900881","https://openalex.org/W4312443924","https://openalex.org/W4400315380","https://openalex.org/W6631190155","https://openalex.org/W6637373629","https://openalex.org/W6638214083","https://openalex.org/W6681435938","https://openalex.org/W6692572551","https://openalex.org/W6737650911","https://openalex.org/W6739706309","https://openalex.org/W6744283545","https://openalex.org/W6745375793","https://openalex.org/W6745558287","https://openalex.org/W6747043858","https://openalex.org/W6747620207","https://openalex.org/W6747991573","https://openalex.org/W6749538638","https://openalex.org/W6751983209","https://openalex.org/W6752314297","https://openalex.org/W6752571064","https://openalex.org/W6752663585","https://openalex.org/W6753170497","https://openalex.org/W6753892653","https://openalex.org/W6755501240","https://openalex.org/W6757030017","https://openalex.org/W6757751764","https://openalex.org/W6757817989","https://openalex.org/W6758207765","https://openalex.org/W6758442962","https://openalex.org/W6762287338","https://openalex.org/W6784164614","https://openalex.org/W6784591795","https://openalex.org/W6786762364","https://openalex.org/W6788135285","https://openalex.org/W6793293946","https://openalex.org/W6803727029","https://openalex.org/W6805973258","https://openalex.org/W6809964999","https://openalex.org/W6838487405","https://openalex.org/W6838504867","https://openalex.org/W7006883545"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W3162204513","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W2033914206","https://openalex.org/W2042327336"],"abstract_inverted_index":{"AdamW":[0,80,91,136,165],"modifies":[1],"Adam":[2,62,65,87,118,148,171],"by":[3,109],"adding":[4],"a":[5,96,105,139],"decoupled":[6,20,110,190],"weight":[7,21,111,191],"decay":[8,10,22,192],"to":[9,137,147],"network":[11],"weights":[12],"per":[13],"training":[14],"iteration.":[15],"For":[16],"adaptive":[17],"algorithms,":[18],"this":[19,73],"does":[23],"not":[24],"affect":[25],"specific":[26],"optimization":[27,38],"steps,":[28],"and":[29,44,58,63,81,88,104,119,127,149,151,172],"differs":[30],"from":[31,174],"the":[32,42,77,175,182,187],"widely":[33],"used":[34],"l<sub>2</sub>-regularizer":[35],"which":[36],"changes":[37],"steps":[39],"via":[40],"changing":[41],"first-":[43],"second-order":[45],"gradient":[46,133],"moments.":[47],"Despite":[48],"its":[49,55,83],"great":[50],"practical":[51],"success,":[52],"for":[53,158,181],"AdamW,":[54],"convergence":[56,78],"behavior":[57],"generalization":[59,84,168],"improvement":[60],"over":[61,86],"l<sub>2</sub>-regularized":[64],"(":[66],"l<sub>2</sub>-Adam)":[67],"remain":[68],"absent":[69],"yet.":[70],"To":[71],"solve":[72],"issue,":[74],"we":[75,130,162],"prove":[76,163],"of":[79,135,189],"justify":[82],"advantages":[85],"l<sub>2</sub>-Adam.":[89,120],"Specifically,":[90],"provably":[92],"converges":[93],"but":[94],"minimizes":[95],"dynamically":[97],"regularized":[98],"loss":[99,103],"that":[100,164],"combines":[101],"vanilla":[102],"dynamical":[106],"regularization":[107],"induced":[108],"decay,":[112],"thus":[113],"yielding":[114],"different":[115],"behaviors":[116],"with":[117],"Moreover,":[121],"on":[122],"both":[123],"general":[124],"nonconvex":[125],"problems":[126],"P\u0141-conditioned":[128],"problems,":[129],"establish":[131],"stochastic":[132],"complexity":[134,143],"find":[138],"stationary":[140],"point.":[141],"Such":[142],"is":[144],"also":[145],"applicable":[146],"l<sub>2</sub>-Adam,":[150],"improves":[152],"their":[153],"previously":[154],"known":[155],"complexity,":[156],"especially":[157],"over-parametrized":[159],"networks.":[160],"Besides,":[161],"enjoys":[166],"smaller":[167],"errors":[169],"than":[170],"l<sub>2</sub>-Adam":[173],"Bayesian":[176],"posterior":[177],"aspect.":[178],"This":[179],"result,":[180],"first":[183],"time,":[184],"explicitly":[185],"reveals":[186],"benefits":[188],"in":[193],"AdamW.":[194],"Experimental":[195],"results":[196],"validate":[197],"our":[198],"theory.":[199]},"counts_by_year":[{"year":2026,"cited_by_count":40},{"year":2025,"cited_by_count":131},{"year":2024,"cited_by_count":9}],"updated_date":"2026-04-14T08:04:32.555800","created_date":"2025-10-10T00:00:00"}