{"id":"https://openalex.org/W7222551925","doi":"https://doi.org/10.48550/arxiv.2610.21026","title":"On the Last-Iterate Convergence of Clipped Gradient Methods","display_name":"On the Last-Iterate Convergence of Clipped Gradient Methods","publication_year":2026,"publication_date":"2026-10-14","ids":{"openalex":"https://openalex.org/W7222551925","doi":"https://doi.org/10.48550/arxiv.2610.21026"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.21026","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.21026","pdf_url":"https://arxiv.org/pdf/2610.21026","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/2610.21026","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5158867154","display_name":"Zijian Liu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I57206974","display_name":"New York University","ror":"https://ror.org/0190ak572","country_code":"US","type":"education","lineage":["https://openalex.org/I57206974"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Liu, Zijian","raw_affiliation_strings":["New York University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"New York University","institution_ids":["https://openalex.org/I57206974"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5158867154"],"corresponding_institution_ids":["https://openalex.org/I57206974"],"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":0.9994999766349792,"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":0.9994999766349792,"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/T10249","display_name":"Distributed Control Multi-Agent Systems","score":0.0,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems 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/T10545","display_name":"Optimization and Variational Analysis","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/last-iterate-convergence","display_name":"last-iterate convergence","score":0.9459999799728394},{"id":"https://openalex.org/keywords/heavy-tailed-noise","display_name":"heavy-tailed noise","score":0.9100000262260437},{"id":"https://openalex.org/keywords/stochastic-convex-optimization","display_name":"stochastic convex optimization","score":0.7319999933242798},{"id":"https://openalex.org/keywords/strongly-convex-optimization","display_name":"strongly convex optimization","score":0.7360000014305115},{"id":"https://openalex.org/keywords/gradient-clipping","display_name":"gradient clipping","score":0.7409999966621399},{"id":"https://openalex.org/keywords/first-order-methods","display_name":"first-order methods","score":0.6110000014305115}],"concepts":[{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.6879000067710876},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.6333000063896179},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.5950000286102295},{"id":"https://openalex.org/C2778775528","wikidata":"https://www.wikidata.org/wiki/Q5135432","display_name":"Closing (real estate)","level":2,"score":0.5444999933242798},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.5346999764442444},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.4977000057697296},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.46239998936653137},{"id":"https://openalex.org/C112680207","wikidata":"https://www.wikidata.org/wiki/Q714886","display_name":"Regular polygon","level":2,"score":0.4593999981880188},{"id":"https://openalex.org/C189950617","wikidata":"https://www.wikidata.org/wiki/Q937228","display_name":"Property (philosophy)","level":2,"score":0.4578999876976013},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.4528999924659729},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.45010000467300415},{"id":"https://openalex.org/C115680565","wikidata":"https://www.wikidata.org/wiki/Q5977448","display_name":"Gradient method","level":2,"score":0.41370001435279846},{"id":"https://openalex.org/C95763700","wikidata":"https://www.wikidata.org/wiki/Q578985","display_name":"Convergence of random variables","level":3,"score":0.3959999978542328},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.3779999911785126},{"id":"https://openalex.org/C145446738","wikidata":"https://www.wikidata.org/wiki/Q319913","display_name":"Convex function","level":3,"score":0.3650999963283539},{"id":"https://openalex.org/C75235859","wikidata":"https://www.wikidata.org/wiki/Q582659","display_name":"Exponential growth","level":2,"score":0.3440999984741211},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.32499998807907104},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3244999945163727},{"id":"https://openalex.org/C57869625","wikidata":"https://www.wikidata.org/wiki/Q1783502","display_name":"Rate of convergence","level":3,"score":0.31949999928474426},{"id":"https://openalex.org/C8272713","wikidata":"https://www.wikidata.org/wiki/Q176737","display_name":"Stochastic process","level":2,"score":0.3034000098705292},{"id":"https://openalex.org/C55479107","wikidata":"https://www.wikidata.org/wiki/Q97663916","display_name":"Stochastic approximation","level":3,"score":0.273499995470047},{"id":"https://openalex.org/C10494615","wikidata":"https://www.wikidata.org/wiki/Q17086765","display_name":"Proximal Gradient Methods","level":4,"score":0.267300009727478},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.258899986743927}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.21026","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.21026","pdf_url":"https://arxiv.org/pdf/2610.21026","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.2610.21026","doi":"https://doi.org/10.48550/arxiv.2610.21026","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.21026","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:2610.21026","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.21026","pdf_url":"https://arxiv.org/pdf/2610.21026","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/W7222551925.pdf","grobid_xml":"https://content.openalex.org/works/W7222551925.grobid-xml"},"referenced_works_count":10,"referenced_works":["https://openalex.org/W2083459869","https://openalex.org/W2112269233","https://openalex.org/W2436219157","https://openalex.org/W2551284144","https://openalex.org/W2940467104","https://openalex.org/W4360826912","https://openalex.org/W4403820306","https://openalex.org/W4414393740","https://openalex.org/W4416837024","https://openalex.org/W6948236204"],"related_works":["https://openalex.org/W7221794827","https://openalex.org/W4389814389","https://openalex.org/W4405903307","https://openalex.org/W3026800838","https://openalex.org/W4392646840","https://openalex.org/W4320721098","https://openalex.org/W7223094074","https://openalex.org/W4403820306","https://openalex.org/W7154618849","https://openalex.org/W6942924577"],"abstract_inverted_index":{"In":[0,141],"stochastic":[1],"(strongly)":[2],"convex":[3],"optimization,":[4],"establishing":[5],"last-iterate":[6,55,135,149],"convergence":[7,136],"guarantees":[8],"for":[9,154],"first-order":[10],"methods":[11],"has":[12],"recently":[13],"garnered":[14],"increasing":[15],"attention,":[16],"as":[17],"this":[18,142],"not":[19],"only":[20],"deepens":[21],"the":[22,42,58,73,83,108,146,164],"theoretical":[23],"understanding":[24],"of":[25,41,110],"these":[26],"algorithms":[27],"but":[28],"also":[29],"aligns":[30],"more":[31,84],"closely":[32],"with":[33,91,122],"practice.":[34],"Notably,":[35],"people":[36,113],"have":[37,62,114],"demonstrated":[38],"that":[39,119],"one":[40],"most":[43],"famous,":[44],"simple,":[45],"and":[46],"popular":[47],"methods,":[48],"Stochastic":[49],"Gradient":[50],"Descent":[51],"($\\mathtt{SGD}$),":[52],"provably":[53],"achieves":[54],"convergence.":[56],"However,":[57],"existing":[59],"results":[60],"may":[61],"limited":[63],"applicability,":[64],"since":[65],"they":[66],"typically":[67],"rely":[68],"on":[69,72],"strong":[70],"conditions":[71],"gradient":[74,126],"noise":[75],"(e.g.,":[76],"an":[77,117],"exponentially":[78],"decaying":[79],"tail)":[80],"rather":[81],"than":[82],"realistic":[85],"heavy-tailed":[86,92,111,157],"noise.":[87],"Unfortunately,":[88],"when":[89],"faced":[90],"noise,":[93,112,158],"$\\mathtt{SGD}$":[94,121],"is":[95],"known":[96],"to":[97,104],"exhibit":[98],"undesirable":[99],"behavior":[100],"or":[101],"even":[102],"fail":[103],"converge.":[105],"To":[106],"address":[107],"challenge":[109],"proposed":[115],"$\\mathtt{Clipped}\\text{-}\\mathtt{SGD}$,":[116],"algorithm":[118],"combines":[120],"a":[123,161],"simple":[124],"mechanism,":[125],"clipping.":[127],"Although":[128],"$\\mathtt{Clipped}\\text{-}\\mathtt{SGD}$":[129,155],"performs":[130],"well":[131],"in":[132,151,163],"practice,":[133],"its":[134],"property":[137],"remains":[138],"highly":[139],"underexplored.":[140],"work,":[143],"we":[144],"prove":[145],"first":[147],"optimal":[148],"rates":[150],"high":[152],"probability":[153],"under":[156],"thereby":[159],"closing":[160],"gap":[162],"literature.":[165]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-17T00:00:00"}