{"id":"https://openalex.org/W7106770756","doi":"https://doi.org/10.48550/arxiv.2511.18006","title":"Understanding Private Learning From Feature Perspective","display_name":"Understanding Private Learning From Feature Perspective","publication_year":2025,"publication_date":"2025-11-22","ids":{"openalex":"https://openalex.org/W7106770756","doi":"https://doi.org/10.48550/arxiv.2511.18006"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2511.18006","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2511.18006","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2511.18006","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Ding, Meng","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Ding, Meng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Lei, Mingxi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lei, Mingxi","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Fu, Shaopeng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Fu, Shaopeng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Wang, Shaowei","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Shaowei","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Wang, Di","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Di","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Xu, Jinhui","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Jinhui","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":6,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9427000284194946,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9427000284194946,"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":0.020099999383091927,"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/T10237","display_name":"Cryptography and Data Security","score":0.007000000216066837,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.7444000244140625},{"id":"https://openalex.org/keywords/memorization","display_name":"Memorization","score":0.637499988079071},{"id":"https://openalex.org/keywords/generalization","display_name":"Generalization","score":0.5734999775886536},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.5313000082969666},{"id":"https://openalex.org/keywords/feature-learning","display_name":"Feature learning","score":0.4932999908924103},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.47609999775886536},{"id":"https://openalex.org/keywords/private-information-retrieval","display_name":"Private information retrieval","score":0.39419999718666077},{"id":"https://openalex.org/keywords/supervised-learning","display_name":"Supervised learning","score":0.3700000047683716},{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"Stochastic gradient descent","score":0.3675999939441681}],"concepts":[{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.7444000244140625},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6689000129699707},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6499999761581421},{"id":"https://openalex.org/C30038468","wikidata":"https://www.wikidata.org/wiki/Q4354775","display_name":"Memorization","level":2,"score":0.637499988079071},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.6258000135421753},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.5734999775886536},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.5313000082969666},{"id":"https://openalex.org/C59404180","wikidata":"https://www.wikidata.org/wiki/Q17013334","display_name":"Feature learning","level":2,"score":0.4932999908924103},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.47609999775886536},{"id":"https://openalex.org/C99221444","wikidata":"https://www.wikidata.org/wiki/Q1532069","display_name":"Private information retrieval","level":2,"score":0.39419999718666077},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.3700000047683716},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.3675999939441681},{"id":"https://openalex.org/C160920958","wikidata":"https://www.wikidata.org/wiki/Q7662746","display_name":"Synthetic data","level":2,"score":0.3598000109195709},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.3361000120639801},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.31690001487731934},{"id":"https://openalex.org/C52622490","wikidata":"https://www.wikidata.org/wiki/Q1026626","display_name":"Feature extraction","level":2,"score":0.2996000051498413},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2890999913215637},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.28859999775886536},{"id":"https://openalex.org/C24138899","wikidata":"https://www.wikidata.org/wiki/Q17141258","display_name":"Instance-based learning","level":3,"score":0.27399998903274536},{"id":"https://openalex.org/C120936955","wikidata":"https://www.wikidata.org/wiki/Q2155640","display_name":"Empirical research","level":2,"score":0.26179999113082886},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.26100000739097595},{"id":"https://openalex.org/C83665646","wikidata":"https://www.wikidata.org/wiki/Q42139305","display_name":"Feature vector","level":2,"score":0.26010000705718994},{"id":"https://openalex.org/C34585555","wikidata":"https://www.wikidata.org/wiki/Q1368723","display_name":"Learning curve","level":2,"score":0.257999986410141},{"id":"https://openalex.org/C2781170535","wikidata":"https://www.wikidata.org/wiki/Q30587856","display_name":"Noisy data","level":2,"score":0.2526000142097473},{"id":"https://openalex.org/C150899416","wikidata":"https://www.wikidata.org/wiki/Q1820378","display_name":"Transfer of learning","level":2,"score":0.2517000138759613},{"id":"https://openalex.org/C90119067","wikidata":"https://www.wikidata.org/wiki/Q43260","display_name":"Polynomial","level":2,"score":0.25040000677108765}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2511.18006","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2511.18006","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2511.18006","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2511.18006","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"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":{"Differentially":[0],"private":[1,41,54,111,123,151,167],"Stochastic":[2],"Gradient":[3],"Descent":[4],"(DP-SGD)":[5],"has":[6],"become":[7],"integral":[8],"to":[9,30,52,133,154,176],"privacy-preserving":[10],"machine":[11],"learning,":[12,145,152],"ensuring":[13],"robust":[14],"privacy":[15],"guarantees":[16],"in":[17,40,87,110,143,150],"sensitive":[18],"domains.":[19],"Despite":[20],"notable":[21],"empirical":[22],"advances":[23],"leveraging":[24],"features":[25],"from":[26,67],"non-private,":[27],"pre-trained":[28],"models":[29],"enhance":[31],"DP-SGD":[32],"training,":[33,135],"a":[34,57,80,92,127],"theoretical":[35,50,188],"understanding":[36],"of":[37,166,173],"feature":[38,58,75,103,174],"dynamics":[39],"learning":[42,59,105,125,168],"remains":[43],"underexplored.":[44],"This":[45],"paper":[46],"presents":[47],"the":[48,63,88,164,171],"first":[49],"framework":[51],"analyze":[53],"training":[55,112,159],"through":[56],"perspective.":[60],"Building":[61],"on":[62,180],"multi-patch":[64],"data":[65,107,139],"structure":[66],"prior":[68],"work,":[69],"our":[70,187],"analysis":[71],"distinguishes":[72],"between":[73],"label-dependent":[74],"signals":[76],"and":[77,106,136,169,182],"label-independent":[78],"noise,":[79],"critical":[81],"aspect":[82],"overlooked":[83],"by":[84],"existing":[85],"analyses":[86],"DP":[89],"community.":[90],"Employing":[91],"two-layer":[93],"CNN":[94],"with":[95],"polynomial":[96],"ReLU":[97],"activation,":[98],"we":[99],"theoretically":[100],"characterize":[101],"both":[102],"signal":[104,124],"noise":[108,140],"memorization":[109,141],"via":[113],"noisy":[114],"gradient":[115],"descent.":[116],"Our":[117,161],"findings":[118,162],"reveal":[119],"that":[120],"(1)":[121],"Effective":[122],"requires":[126],"higher":[128],"signal-to-noise":[129],"ratio":[130],"(SNR)":[131],"compared":[132],"non-private":[134,144],"(2)":[137],"When":[138],"occurs":[142],"it":[146],"will":[147],"also":[148,185],"occur":[149],"leading":[153],"poor":[155],"generalization":[156],"despite":[157],"small":[158],"loss.":[160],"highlight":[163],"challenges":[165],"prove":[170],"benefit":[172],"enhancement":[175],"improve":[177],"SNR.":[178],"Experiments":[179],"synthetic":[181],"real-world":[183],"datasets":[184],"validate":[186],"findings.":[189]},"counts_by_year":[],"updated_date":"2025-11-27T01:16:37.896743","created_date":"2025-11-27T00:00:00"}