{"id":"https://openalex.org/W7164936627","doi":"https://doi.org/10.48550/arxiv.2606.15832","title":"SILAGE: Memory-Efficient, Full-Gradient-Free Nonconvex Optimization for Nested Finite Sums","display_name":"SILAGE: Memory-Efficient, Full-Gradient-Free Nonconvex Optimization for Nested Finite Sums","publication_year":2026,"publication_date":"2026-06-14","ids":{"openalex":"https://openalex.org/W7164936627","doi":"https://doi.org/10.48550/arxiv.2606.15832"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.15832","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.15832","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.15832","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5102027815","display_name":"Igor Sokolov","orcid":"https://orcid.org/0000-0002-2338-0187"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sokolov, Igor","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138732500","display_name":"Laurent Condat","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Condat, Laurent","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138705081","display_name":"Peter Richt\u00e1rik","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Richt\u00e1rik, Peter","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":false,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9541000127792358,"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":0.9541000127792358,"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/T10136","display_name":"Statistical Methods and Inference","score":0.00930000003427267,"subfield":{"id":"https://openalex.org/subfields/2613","display_name":"Statistics and Probability"},"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/T10500","display_name":"Sparse and Compressive Sensing Techniques","score":0.008999999612569809,"subfield":{"id":"https://openalex.org/subfields/2206","display_name":"Computational Mechanics"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/lipschitz-continuity","display_name":"Lipschitz continuity","score":0.6155999898910522},{"id":"https://openalex.org/keywords/oracle","display_name":"Oracle","score":0.5659000277519226},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.5346999764442444},{"id":"https://openalex.org/keywords/minification","display_name":"Minification","score":0.517300009727478},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.4187999963760376},{"id":"https://openalex.org/keywords/nested-loop-join","display_name":"Nested loop join","score":0.349700003862381},{"id":"https://openalex.org/keywords/heuristics","display_name":"Heuristics","score":0.3352000117301941},{"id":"https://openalex.org/keywords/gradient-descent","display_name":"Gradient descent","score":0.3163999915122986}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6195999979972839},{"id":"https://openalex.org/C22324862","wikidata":"https://www.wikidata.org/wiki/Q652707","display_name":"Lipschitz continuity","level":2,"score":0.6155999898910522},{"id":"https://openalex.org/C55166926","wikidata":"https://www.wikidata.org/wiki/Q2892946","display_name":"Oracle","level":2,"score":0.5659000277519226},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.5346999764442444},{"id":"https://openalex.org/C147764199","wikidata":"https://www.wikidata.org/wiki/Q6865248","display_name":"Minification","level":2,"score":0.517300009727478},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.4187999963760376},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.41110000014305115},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3634999990463257},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3598000109195709},{"id":"https://openalex.org/C1306188","wikidata":"https://www.wikidata.org/wiki/Q4060687","display_name":"Nested loop join","level":2,"score":0.349700003862381},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.3352000117301941},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.3163999915122986},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.3027999997138977},{"id":"https://openalex.org/C164752517","wikidata":"https://www.wikidata.org/wiki/Q5570875","display_name":"Global optimization","level":2,"score":0.29739999771118164},{"id":"https://openalex.org/C2781311116","wikidata":"https://www.wikidata.org/wiki/Q83306","display_name":"Group (periodic table)","level":2,"score":0.2973000109195709},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.289000004529953},{"id":"https://openalex.org/C145446738","wikidata":"https://www.wikidata.org/wiki/Q319913","display_name":"Convex function","level":3,"score":0.27320000529289246},{"id":"https://openalex.org/C75553542","wikidata":"https://www.wikidata.org/wiki/Q178161","display_name":"A priori and a posteriori","level":2,"score":0.27140000462532043},{"id":"https://openalex.org/C157972887","wikidata":"https://www.wikidata.org/wiki/Q463359","display_name":"Convex optimization","level":3,"score":0.26460000872612},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C112680207","wikidata":"https://www.wikidata.org/wiki/Q714886","display_name":"Regular polygon","level":2,"score":0.258899986743927},{"id":"https://openalex.org/C132943942","wikidata":"https://www.wikidata.org/wiki/Q2562511","display_name":"Footprint","level":2,"score":0.25519999861717224},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2542000114917755}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.15832","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.15832","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.15832","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.15832","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,"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"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Empirical":[0],"risk":[1],"minimization":[2],"on":[3],"massive":[4],"datasets":[5],"naturally":[6],"exhibits":[7],"a":[8,94,106,146],"nested":[9,167],"double":[10],"finite-sum":[11],"structure,":[12,118],"where":[13],"$N=nm$":[14],"total":[15],"samples":[16],"are":[17,73],"logically":[18],"or":[19,35],"physically":[20],"partitioned":[21],"into":[22],"$n$":[23],"blocks":[24],"of":[25],"size":[26],"$m$":[27],"(e.g.,":[28],"in":[29,54,182],"pooled":[30],"data":[31,164],"silos,":[32],"out-of-core":[33],"learning,":[34],"deliberate":[36],"stratification).":[37],"While":[38],"variance-reduced":[39,107],"methods":[40],"achieve":[41],"optimal":[42],"oracle":[43],"complexities":[44],"for":[45,97],"nonconvex":[46],"objectives,":[47],"they":[48],"suffer":[49],"from":[50],"severe":[51],"scaling":[52],"bottlenecks":[53],"this":[55,101,111],"centralized":[56],"regime.":[57],"Recursive":[58],"estimators,":[59],"such":[60,79,83],"as":[61,80],"PAGE,":[62],"require":[63,86],"periodic":[64,121],"global":[65,122],"full-gradient":[66,123],"refreshes":[67,84,124],"over":[68,125],"all":[69,126],"$nm$":[70,127],"samples,":[71],"which":[72],"computationally":[74],"expensive.":[75],"Conversely,":[76],"single-loop":[77],"methods,":[78],"SILVER,":[81],"avoid":[82],"but":[85],"an":[87],"impractical":[88],"$\\mathcal{O}(nm)$":[89],"memory":[90],"footprint":[91],"to":[92,161],"store":[93],"control":[95],"variate":[96],"every":[98],"sample.":[99],"In":[100],"paper,":[102],"we":[103,144],"propose":[104],"SILAGE,":[105],"algorithm":[108],"that":[109,150],"addresses":[110],"trade-off.":[112],"By":[113],"actively":[114],"exploiting":[115],"the":[116,162],"double-sum":[117],"SILAGE":[119],"eliminates":[120],"components":[128],"(evaluating":[129],"at":[130],"most":[131],"one":[132],"local":[133],"group":[134],"gradient":[135],"per":[136],"iteration)":[137],"while":[138],"requiring":[139],"only":[140],"$\\mathcal{O}(n)$":[141],"memory.":[142],"Furthermore,":[143],"provide":[145],"tight":[147],"convergence":[148],"analysis":[149],"avoids":[151],"pessimistic":[152],"worst-case":[153],"Lipschitz":[154],"constants.":[155],"Instead,":[156],"SILAGE's":[157],"complexity":[158],"natively":[159],"adapts":[160],"underlying":[163],"geometry":[165],"via":[166],"functional":[168],"similarities:":[169],"across-group":[170],"($\u03b4_1$)":[171],"and":[172],"within-group":[173],"($\u03b4_2$)":[174],"heterogeneity.":[175],"Our":[176],"results":[177],"improve":[178],"existing":[179],"state-of-the-art":[180],"bounds":[181],"several":[183],"practically":[184],"relevant":[185],"regimes.":[186]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-06-17T00:00:00"}