{"id":"https://openalex.org/W3213048841","doi":"https://doi.org/10.48550/arxiv.2111.14069","title":"Escape saddle points by a simple gradient-descent based algorithm","display_name":"Escape saddle points by a simple gradient-descent based algorithm","publication_year":2021,"publication_date":"2021-11-28","ids":{"openalex":"https://openalex.org/W3213048841","doi":"https://doi.org/10.48550/arxiv.2111.14069","mag":"3213048841"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2111.14069","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2111.14069","pdf_url":"https://arxiv.org/pdf/2111.14069","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2111.14069","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Zhang, Chenyi","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Chenyi","raw_affiliation_strings":["Institute for Interdisciplinary Information Sciences , Tsinghua University , China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Institute for Interdisciplinary Information Sciences , Tsinghua University , China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101509600","display_name":"Tongyang Li","orcid":"https://orcid.org/0000-0002-0338-413X","observed_orcids":["https://orcid.org/0000-0002-0338-413X"]},"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":"Li, Tongyang","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":5,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":"34","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":0.9824000000953674,"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":0.9824000000953674,"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/T10963","display_name":"Advanced Optimization Algorithms Research","score":0.014399999752640724,"subfield":{"id":"https://openalex.org/subfields/2612","display_name":"Numerical Analysis"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10545","display_name":"Optimization and Variational Analysis","score":0.000699999975040555,"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/saddle-point-escape","display_name":"saddle point escape","score":0.8799999952316284},{"id":"https://openalex.org/keywords/nonconvex-optimization","display_name":"non-convex optimization","score":0.9620000123977661},{"id":"https://openalex.org/keywords/second-order-stationary-points","display_name":"second-order stationary points","score":0.902999997138977},{"id":"https://openalex.org/keywords/gradient-based-optimization","display_name":"gradient-based optimization","score":0.621999979019165},{"id":"https://openalex.org/keywords/negative-curvature","display_name":"negative curvature","score":0.906000018119812},{"id":"https://openalex.org/keywords/stochastic-optimization","display_name":"stochastic optimization","score":0.6710000038146973},{"id":"https://openalex.org/keywords/smooth-functions","display_name":"smooth functions","score":0.4399999976158142}],"concepts":[{"id":"https://openalex.org/C2681867","wikidata":"https://www.wikidata.org/wiki/Q690935","display_name":"Saddle point","level":2,"score":0.808668315410614},{"id":"https://openalex.org/C203616005","wikidata":"https://www.wikidata.org/wiki/Q620495","display_name":"Hessian matrix","level":2,"score":0.7496766448020935},{"id":"https://openalex.org/C189237950","wikidata":"https://www.wikidata.org/wiki/Q2500758","display_name":"Stationary point","level":2,"score":0.6471837162971497},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.6161624789237976},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.6029178500175476},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.5803167223930359},{"id":"https://openalex.org/C182306322","wikidata":"https://www.wikidata.org/wiki/Q1779371","display_name":"Order (exchange)","level":2,"score":0.5741205811500549},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4997107982635498},{"id":"https://openalex.org/C63553672","wikidata":"https://www.wikidata.org/wiki/Q581168","display_name":"Binary logarithm","level":2,"score":0.4802829921245575},{"id":"https://openalex.org/C311688","wikidata":"https://www.wikidata.org/wiki/Q2393193","display_name":"Time complexity","level":2,"score":0.4766809940338135},{"id":"https://openalex.org/C2780586882","wikidata":"https://www.wikidata.org/wiki/Q7520643","display_name":"Simple (philosophy)","level":2,"score":0.4625723958015442},{"id":"https://openalex.org/C90119067","wikidata":"https://www.wikidata.org/wiki/Q43260","display_name":"Polynomial","level":2,"score":0.4455733597278595},{"id":"https://openalex.org/C195065555","wikidata":"https://www.wikidata.org/wiki/Q214881","display_name":"Curvature","level":2,"score":0.43575263023376465},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.4116893708705902},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.2267611026763916},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.18905293941497803},{"id":"https://openalex.org/C28826006","wikidata":"https://www.wikidata.org/wiki/Q33521","display_name":"Applied mathematics","level":1,"score":0.18156543374061584},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.15693065524101257},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.0},{"id":"https://openalex.org/C10138342","wikidata":"https://www.wikidata.org/wiki/Q43015","display_name":"Finance","level":1,"score":0.0},{"id":"https://openalex.org/C111472728","wikidata":"https://www.wikidata.org/wiki/Q9471","display_name":"Epistemology","level":1,"score":0.0},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.0}],"mesh":[],"locations_count":3,"locations":[{"id":"pmh:oai:arXiv.org:2111.14069","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2111.14069","pdf_url":"https://arxiv.org/pdf/2111.14069","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2111.14069","doi":"https://doi.org/10.48550/arxiv.2111.14069","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2111.14069","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":null,"raw_source_name":null,"raw_type":"Preprint"},{"id":"mag:3213048841","doi":null,"is_oa":false,"landing_page_url":"https://openreview.net/forum?id=lEf52hTHq0Q","pdf_url":null,"source":{"id":"https://openalex.org/S4363606243","display_name":"neural information processing systems","issn_l":"1049-5258","issn":["1049-5258"],"is_oa":false,"is_in_doaj":false,"is_core":true,"listed_in":["cwts-core","norway-2"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":"Neural Information Processing Systems","raw_type":null}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2111.14069","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2111.14069","pdf_url":"https://arxiv.org/pdf/2111.14069","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G107522343","display_name":"PFCQC: STAQ: Software-Tailored Architecture for Quantum co-design","funder_award_id":"1818914","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G6552274349","display_name":null,"funder_award_id":"1818914","funder_id":"https://openalex.org/F4320315121","funder_display_name":"Samsung Advanced Institute of Technology"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320315121","display_name":"Samsung Advanced Institute of Technology","ror":"https://ror.org/04axnyp10"},{"id":"https://openalex.org/F4320332195","display_name":"Samsung","ror":"https://ror.org/04w3jy968"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2911876570","https://openalex.org/W2769147973","https://openalex.org/W2474090883","https://openalex.org/W3020927592","https://openalex.org/W3120456000","https://openalex.org/W2495922485","https://openalex.org/W3111569753","https://openalex.org/W3111986636","https://openalex.org/W2891466145","https://openalex.org/W1984402599","https://openalex.org/W3171826361","https://openalex.org/W2935777426","https://openalex.org/W1989566487","https://openalex.org/W2073312178","https://openalex.org/W3020853991","https://openalex.org/W1995989058","https://openalex.org/W3105660007","https://openalex.org/W2914300050","https://openalex.org/W2937833101","https://openalex.org/W2968762515"],"abstract_inverted_index":{"Escaping":[0],"saddle":[1,112],"points":[2,113],"is":[3,57,93],"a":[4,16,23,98],"central":[5],"research":[6],"topic":[7],"in":[8,34,60,69,85,119],"nonconvex":[9],"optimization.":[10],"In":[11],"this":[12],"paper,":[13],"we":[14,130],"propose":[15],"simple":[17],"gradient-based":[18],"algorithm":[19,56,78],"such":[20],"that":[21,135],"for":[22],"smooth":[24],"function":[25],"$f\\colon\\mathbb{R}^n\\to\\mathbb{R}$,":[26],"it":[27],"outputs":[28,79],"an":[29,80,94],"$\u03b5$-approximate":[30,81],"second-order":[31,82],"stationary":[32,83],"point":[33,84],"$\\tilde{O}(\\log":[35],"n/\u03b5^{1.75})$":[36],"iterations.":[37,88],"Compared":[38],"to":[39,123],"the":[40,74,116,124],"previous":[41],"state-of-the-art":[42],"algorithms":[43],"by":[44],"Jin":[45],"et":[46],"al.":[47],"with":[48],"$\\tilde{O}((\\log":[49,52,86],"n)^{4}/\u03b5^{2})$":[50],"or":[51],"n)^{6}/\u03b5^{1.75})$":[53],"iterations,":[54],"our":[55,77,90,137],"polynomially":[58],"better":[59],"terms":[61,70],"of":[62,71,96],"$\\log":[63,120],"n$":[64,121],"and":[65,114],"matches":[66],"their":[67],"complexities":[68],"$1/\u03b5$.":[72],"For":[73],"stochastic":[75],"setting,":[76],"n)^{2}/\u03b5^{4})$":[87],"Technically,":[89],"main":[91],"contribution":[92],"idea":[95],"implementing":[97],"robust":[99],"Hessian":[100],"power":[101],"method":[102],"using":[103],"only":[104],"gradients,":[105],"which":[106],"can":[107],"find":[108],"negative":[109],"curvature":[110],"near":[111],"achieve":[115],"polynomial":[117],"speedup":[118],"compared":[122],"perturbed":[125],"gradient":[126],"descent":[127],"methods.":[128],"Finally,":[129],"also":[131],"perform":[132],"numerical":[133],"experiments":[134],"support":[136],"results.":[138]},"counts_by_year":[{"year":2024,"cited_by_count":2},{"year":2023,"cited_by_count":2},{"year":2022,"cited_by_count":1}],"updated_date":"2026-10-16T09:59:22.235173","created_date":"2025-10-10T00:00:00"}