{"id":"https://openalex.org/W3135897355","doi":null,"title":"More data or more parameters? Investigating the effect of data structure on generalization.","display_name":"More data or more parameters? Investigating the effect of data structure on generalization.","publication_year":2021,"publication_date":"2021-03-09","ids":{"openalex":"https://openalex.org/W3135897355","mag":"3135897355"},"language":"en","primary_location":{"id":"mag:3135897355","doi":null,"is_oa":true,"landing_page_url":"http://export.arxiv.org/pdf/2103.05524v1","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":[],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"http://export.arxiv.org/pdf/2103.05524v1","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5077817630","display_name":"St\u00e9phane d\u2019Ascoli","orcid":"https://orcid.org/0000-0002-3131-3371","observed_orcids":["https://orcid.org/0000-0002-3131-3371"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"St\u00e9phane d\u2019Ascoli","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0002-3131-3371","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5001522727","display_name":"Marylou Gabri\u00e9","orcid":"https://orcid.org/0000-0002-5989-1018","observed_orcids":["https://orcid.org/0000-0002-5989-1018"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Marylou Gabri\u00e9","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0002-5989-1018","affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071931582","display_name":"Levent Sagun","orcid":"https://orcid.org/0000-0001-5403-4124","observed_orcids":["https://orcid.org/0000-0001-5403-4124"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Levent Sagun","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0001-5403-4124","affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5002294199","display_name":"Giulio Biroli","orcid":"https://orcid.org/0000-0003-4115-2644","observed_orcids":["https://orcid.org/0000-0003-4115-2644"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Giulio Biroli","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0000-0003-4115-2644","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":2,"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":0.6226000189781189,"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":0.6226000189781189,"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/T10320","display_name":"Neural Networks and Applications","score":0.25119999051094055,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.038100000470876694,"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/generalization","display_name":"generalization","score":0.8050000071525574},{"id":"https://openalex.org/keywords/data-structures","display_name":"data structures","score":0.8080000281333923},{"id":"https://openalex.org/keywords/overparameterization","display_name":"overparameterization","score":0.8659999966621399},{"id":"https://openalex.org/keywords/teacher-student-models","display_name":"teacher-student models","score":0.7820000052452087},{"id":"https://openalex.org/keywords/high-dimensional-limit","display_name":"high-dimensional limit","score":0.7070000171661377},{"id":"https://openalex.org/keywords/phase-space","display_name":"phase space","score":0.5109999775886536},{"id":"https://openalex.org/keywords/noisy-labels","display_name":"label noise","score":0.8289999961853027},{"id":"https://openalex.org/keywords/low-dimensional-systems","display_name":"low-dimensional systems","score":0.578000009059906},{"id":"https://openalex.org/keywords/classification","display_name":"classification","score":0.8870000243186951},{"id":"https://openalex.org/keywords/fully-connected-neural-network","display_name":"fully connected neural network","score":0.7059999704360962}],"concepts":[{"id":"https://openalex.org/C190502265","wikidata":"https://www.wikidata.org/wiki/Q17069496","display_name":"MNIST database","level":3,"score":0.8138847351074219},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.6839839220046997},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.596176266670227},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5825074911117554},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.5821848511695862},{"id":"https://openalex.org/C16910744","wikidata":"https://www.wikidata.org/wiki/Q7705759","display_name":"Test data","level":2,"score":0.5652047395706177},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.5408139824867249},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5203944444656372},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5036353468894958},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.4955960512161255},{"id":"https://openalex.org/C202887219","wikidata":"https://www.wikidata.org/wiki/Q3895221","display_name":"Parametrization (atmospheric modeling)","level":3,"score":0.4509985148906708},{"id":"https://openalex.org/C2778572836","wikidata":"https://www.wikidata.org/wiki/Q380933","display_name":"Space (punctuation)","level":2,"score":0.43493467569351196},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.42187997698783875},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.2883950471878052},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.08108943700790405},{"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/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.0},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C78458016","wikidata":"https://www.wikidata.org/wiki/Q840400","display_name":"Evolutionary biology","level":1,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C74902906","wikidata":"https://www.wikidata.org/wiki/Q1190858","display_name":"Radiative transfer","level":2,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C120665830","wikidata":"https://www.wikidata.org/wiki/Q14620","display_name":"Optics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"mag:3135897355","doi":null,"is_oa":true,"landing_page_url":"http://export.arxiv.org/pdf/2103.05524v1","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null}],"best_oa_location":{"id":"mag:3135897355","doi":null,"is_oa":true,"landing_page_url":"http://export.arxiv.org/pdf/2103.05524v1","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"arXiv (Cornell University)","raw_type":null},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":47,"referenced_works":["https://openalex.org/W1964862779","https://openalex.org/W2009472032","https://openalex.org/W2160815625","https://openalex.org/W2273223125","https://openalex.org/W2618069152","https://openalex.org/W2763894180","https://openalex.org/W2803439868","https://openalex.org/W2807299122","https://openalex.org/W2808465306","https://openalex.org/W2907047316","https://openalex.org/W2918745211","https://openalex.org/W2919115771","https://openalex.org/W2923764619","https://openalex.org/W2950220847","https://openalex.org/W2959995783","https://openalex.org/W2963518130","https://openalex.org/W2964137095","https://openalex.org/W2970388773","https://openalex.org/W2971363918","https://openalex.org/W2977813877","https://openalex.org/W2994081359","https://openalex.org/W2994588524","https://openalex.org/W2996141621","https://openalex.org/W2996360471","https://openalex.org/W3003297907","https://openalex.org/W3004510491","https://openalex.org/W3006943693","https://openalex.org/W3008127699","https://openalex.org/W3009460743","https://openalex.org/W3025991882","https://openalex.org/W3030163527","https://openalex.org/W3034892535","https://openalex.org/W3034893234","https://openalex.org/W3036147742","https://openalex.org/W3037180580","https://openalex.org/W3046835365","https://openalex.org/W3098362323","https://openalex.org/W3098536204","https://openalex.org/W3100072991","https://openalex.org/W3102351945","https://openalex.org/W3151137625","https://openalex.org/W3158305692","https://openalex.org/W3168158992","https://openalex.org/W3172995164","https://openalex.org/W4287752094","https://openalex.org/W4312426305","https://openalex.org/W4409974857"],"related_works":["https://openalex.org/W3211851805","https://openalex.org/W4205437871","https://openalex.org/W3172995164","https://openalex.org/W2950220847","https://openalex.org/W2967536008","https://openalex.org/W3170953504","https://openalex.org/W4399287387","https://openalex.org/W2566079294","https://openalex.org/W3034612893","https://openalex.org/W2907047316"],"abstract_inverted_index":{"One":[0],"of":[1,5,12,29,38,41,46,55,81,91,143,158,166,176],"the":[2,9,32,39,53,58,66,72,89,100,114,126,138,144,151,159,164,174],"central":[3],"features":[4,68],"deep":[6],"learning":[7],"is":[8,79],"generalization":[59],"abilities":[11],"neural":[13],"networks,":[14],"which":[15,84],"seem":[16],"to":[17,87,99,133],"improve":[18],"relentlessly":[19],"with":[20,97],"over-parametrization.":[21],"In":[22],"this":[23],"work,":[24],"we":[25,108],"investigate":[26],"how":[27,52],"properties":[28],"data":[30,56,78,146],"impact":[31],"test":[33,117,152],"error":[34],"as":[35],"a":[36],"function":[37],"number":[40,45,165,175],"training":[42,47,167,177],"examples":[43,168],"and":[44,94,116,122,140,193],"parameters;":[48],"in":[49,71,125,137],"other":[50],"words,":[51],"structure":[54],"shapes":[57],"phase":[60,160],"space.":[61],"We":[62],"first":[63],"focus":[64],"on":[65,150,191],"random":[67],"model":[69],"trained":[70,190],"teacher-student":[73],"scenario.":[74],"The":[75,129],"synthetic":[76],"input":[77,145],"composed":[80],"independent":[82],"blocks,":[83],"allow":[85],"us":[86,132],"tune":[88],"saliency":[90],"low-dimensional":[92],"structures":[93],"their":[95],"relevance":[96],"respect":[98],"target":[101],"function.":[102],"Using":[103],"methods":[104],"from":[105],"statistical":[106],"physics,":[107],"obtain":[109],"an":[110,156],"analytical":[111,180],"expression":[112],"for":[113,119],"train":[115],"errors":[118],"both":[120],"regression":[121],"classification":[123],"tasks":[124],"high-dimensional":[127],"limit.":[128],"derivation":[130],"allows":[131],"show":[134],"that":[135],"noise":[136],"labels":[139],"strong":[141],"anisotropy":[142],"play":[147],"similar":[148],"roles":[149],"error.":[153],"Both":[154],"promote":[155],"asymmetry":[157],"space":[161],"where":[162],"increasing":[163,173],"improves":[169],"further":[171],"than":[172],"parameters.":[178],"Our":[179],"insights":[181],"are":[182],"confirmed":[183],"by":[184],"numerical":[185],"experiments":[186],"involving":[187],"fully-connected":[188],"networks":[189],"MNIST":[192],"CIFAR10.":[194]},"counts_by_year":[{"year":2022,"cited_by_count":1},{"year":2021,"cited_by_count":1}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2021-03-15T00:00:00"}