{"id":"https://openalex.org/W7216169538","doi":"https://doi.org/10.48550/arxiv.2610.00751","title":"Signal-Noise Factorization Isolates Nuisance Variation into Removable Subspaces","display_name":"Signal-Noise Factorization Isolates Nuisance Variation into Removable Subspaces","publication_year":2026,"publication_date":"2026-09-30","ids":{"openalex":"https://openalex.org/W7216169538","doi":"https://doi.org/10.48550/arxiv.2610.00751"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.00751","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.00751","pdf_url":"https://arxiv.org/pdf/2610.00751","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2610.00751","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5081258430","display_name":"Sakin Kirti","orcid":"https://orcid.org/0000-0003-0752-3464","observed_orcids":["https://orcid.org/0000-0003-0752-3464"]},"institutions":[{"id":"https://openalex.org/I161318765","display_name":"University of California, Los Angeles","ror":"https://ror.org/046rm7j60","country_code":"US","type":"education","lineage":["https://openalex.org/I161318765"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kirti, Sakin","raw_affiliation_strings":["Bioinformatics IDP , Jules Stein Eye Institute University of California , Los Angeles Los Angeles , CA 90095 , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Bioinformatics IDP , Jules Stein Eye Institute University of California , Los Angeles Los Angeles , CA 90095 , USA","institution_ids":["https://openalex.org/I161318765"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5028769863","display_name":"Joel Zylberberg","orcid":"https://orcid.org/0000-0002-8208-5698","observed_orcids":["https://orcid.org/0000-0002-8208-5698"]},"institutions":[{"id":"https://openalex.org/I161318765","display_name":"University of California, Los Angeles","ror":"https://ror.org/046rm7j60","country_code":"US","type":"education","lineage":["https://openalex.org/I161318765"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zylberberg, Joel","raw_affiliation_strings":["Jules Stein Eye Institute University of California , Los Angeles Los Angeles , CA , 90095 , USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Jules Stein Eye Institute University of California , Los Angeles Los Angeles , CA , 90095 , USA","institution_ids":["https://openalex.org/I161318765"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I161318765"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.5228999853134155,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.5228999853134155,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.17599999904632568,"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":0.11379999667406082,"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/representational-geometry","display_name":"representational geometry","score":0.9660000205039978},{"id":"https://openalex.org/keywords/regularization","display_name":"regularization","score":0.5600000023841858},{"id":"https://openalex.org/keywords/nuisance-parameters","display_name":"nuisance parameters","score":0.6679999828338623},{"id":"https://openalex.org/keywords/out-of-distribution-robustness","display_name":"out-of-distribution robustness","score":0.796999990940094},{"id":"https://openalex.org/keywords/image-corruptions","display_name":"image corruptions","score":0.5860000252723694},{"id":"https://openalex.org/keywords/deep-neural-network","display_name":"deep neural network","score":0.656000018119812}],"concepts":[{"id":"https://openalex.org/C5274069","wikidata":"https://www.wikidata.org/wiki/Q2285707","display_name":"Categorical variable","level":2,"score":0.7940999865531921},{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.6478999853134155},{"id":"https://openalex.org/C187834632","wikidata":"https://www.wikidata.org/wiki/Q188804","display_name":"Factorization","level":2,"score":0.6184999942779541},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6129999756813049},{"id":"https://openalex.org/C12362212","wikidata":"https://www.wikidata.org/wiki/Q728435","display_name":"Linear subspace","level":2,"score":0.54339998960495},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5412999987602234},{"id":"https://openalex.org/C2776359362","wikidata":"https://www.wikidata.org/wiki/Q2145286","display_name":"Representation (politics)","level":3,"score":0.5131999850273132},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.47099998593330383},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.4499000012874603},{"id":"https://openalex.org/C42355184","wikidata":"https://www.wikidata.org/wiki/Q1361088","display_name":"Matrix decomposition","level":3,"score":0.4480000138282776},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.43160000443458557},{"id":"https://openalex.org/C2778334786","wikidata":"https://www.wikidata.org/wiki/Q1586270","display_name":"Variation (astronomy)","level":2,"score":0.3490999937057495},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3393000066280365},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.3269999921321869},{"id":"https://openalex.org/C152671427","wikidata":"https://www.wikidata.org/wiki/Q10843505","display_name":"Non-negative matrix factorization","level":4,"score":0.3253999948501587},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.3140999972820282},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.3077999949455261},{"id":"https://openalex.org/C203868755","wikidata":"https://www.wikidata.org/wiki/Q5353562","display_name":"Elastic net regularization","level":3,"score":0.29670000076293945},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.29170000553131104},{"id":"https://openalex.org/C134261354","wikidata":"https://www.wikidata.org/wiki/Q938438","display_name":"Statistical inference","level":2,"score":0.27559998631477356},{"id":"https://openalex.org/C194531419","wikidata":"https://www.wikidata.org/wiki/Q17104825","display_name":"Nuisance parameter","level":3,"score":0.26100000739097595},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.2596000134944916},{"id":"https://openalex.org/C116409475","wikidata":"https://www.wikidata.org/wiki/Q1385056","display_name":"External Data Representation","level":2,"score":0.25850000977516174}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.00751","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.00751","pdf_url":"https://arxiv.org/pdf/2610.00751","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.2610.00751","doi":"https://doi.org/10.48550/arxiv.2610.00751","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.00751","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":"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2610.00751","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.00751","pdf_url":"https://arxiv.org/pdf/2610.00751","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":[],"study_designs":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320306811","display_name":"Research to Prevent Blindness","ror":"https://ror.org/04drjs621"},{"id":"https://openalex.org/F4320309949","display_name":"Canadian Institute for Advanced Research","ror":"https://ror.org/01sdtdd95"},{"id":"https://openalex.org/F4320332161","display_name":"National Institutes of Health","ror":"https://ror.org/01cwqze88"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7216169538.pdf","grobid_xml":"https://content.openalex.org/works/W7216169538.grobid-xml"},"referenced_works_count":25,"referenced_works":["https://openalex.org/W2117756735","https://openalex.org/W2127265926","https://openalex.org/W2518108298","https://openalex.org/W2622263826","https://openalex.org/W2811374795","https://openalex.org/W2898564876","https://openalex.org/W2947451636","https://openalex.org/W2949650786","https://openalex.org/W2953494151","https://openalex.org/W3065974826","https://openalex.org/W3093142758","https://openalex.org/W3134652006","https://openalex.org/W3159913668","https://openalex.org/W3173832704","https://openalex.org/W4225522981","https://openalex.org/W4243334717","https://openalex.org/W4286695273","https://openalex.org/W4317436377","https://openalex.org/W4391473051","https://openalex.org/W4400104233","https://openalex.org/W4405419023","https://openalex.org/W4413019869","https://openalex.org/W7104962630","https://openalex.org/W7127662003","https://openalex.org/W7131383102"],"related_works":["https://openalex.org/W2557570108","https://openalex.org/W4377041915","https://openalex.org/W7137564317","https://openalex.org/W6903718167","https://openalex.org/W3003214030","https://openalex.org/W2946794714","https://openalex.org/W2341308508","https://openalex.org/W2985509595","https://openalex.org/W2951748910","https://openalex.org/W3119051753"],"abstract_inverted_index":{"Recent":[0],"theoretical":[1],"work":[2],"identified":[3],"fundamental":[4],"properties":[5,49],"of":[6,13,180,239],"representation":[7,74],"geometry":[8,75,164],"that":[9,45,149,195,197,209],"shape":[10,73],"inference":[11,215],"ability":[12,23,34],"deep":[14],"neural":[15],"networks.":[16],"These":[17,192],"include":[18],"signal-noise":[19,199],"factorization":[20,31,200],"(SNF),":[21],"the":[22,33,64,109,127,132,142,150,167,181],"to":[24,35,59,68,186],"segregate":[25,36],"signal":[26],"from":[27,159,229],"noise,":[28],"and":[29,38,76,120,177],"signal-signal":[30],"(SSF),":[32],"task-specific":[37],"task-irrelevant":[39],"signals.":[40],"Here,":[41],"we":[42,101,140],"built":[43],"regularizers":[44,58,72,105,196],"reinforce":[46],"these":[47,57],"two":[48],"during":[50],"training.":[51],"We":[52],"compared":[53],"networks":[54,62],"trained":[55],"with":[56,113],"$L_2$-regularized":[60],"baseline":[61],"on":[63,79,108,174,205],"CIFAR-100":[65],"classification":[66],"task":[67],"understand":[69,131],"how":[70,103,220],"our":[71,104],"impact":[77],"performance":[78,91,107,124],"a":[80,187],"well-known":[81],"computer":[82,206],"vision":[83,207],"baseline.":[84],"Enhancing":[85],"SNF":[86,128],"via":[87],"regularization":[88,146],"improved":[89,138],"model":[90,224],"but":[92],"enhancing":[93],"SSF":[94],"did":[95],"not.":[96],"Motivated":[97],"by":[98,134],"biomedical":[99],"applications,":[100],"investigated":[102],"affected":[106],"BloodMNIST":[110],"dataset":[111],"treated":[112],"MedMNIST-C":[114],"corruptions":[115],"at":[116,214],"five":[117],"severity":[118],"levels,":[119],"found":[121],"even":[122],"larger":[123],"gains":[125],"using":[126],"regularizer.":[129],"To":[130],"mechanism":[133],"which":[135],"SNF-regularization":[136],"produces":[137],"performance,":[139],"analyzed":[141],"nuisance":[143,227],"subspaces":[144],"across":[145],"regimes,":[147],"finding":[148],"SNF-regularized":[151],"models":[152],"represent":[153],"noise":[154],"in":[155,190],"distinct":[156],"subspaces,":[157],"separate":[158],"class-relevant":[160],"signal.":[161],"Because":[162],"this":[163],"is":[165,232],"explicit,":[166],"dominant":[168],"corruption-induced":[169],"directions":[170],"can":[171,201],"be":[172],"estimated":[173],"held-out":[175],"data":[176],"projected":[178],"out":[179],"representations.":[182],"This":[183],"manipulation":[184],"led":[185],"substantial":[188,203],"gain":[189],"accuracy.":[191],"results":[193],"show":[194],"enforce":[198],"produce":[202],"improvements":[204],"tasks":[208],"contain":[210],"out-of-distribution":[211],"image":[212],"distortions":[213],"time.":[216],"They":[217],"also":[218],"highlight":[219],"shaping":[221],"representations":[222,238],"affects":[223],"performance:":[225],"isolating":[226],"variables":[228],"categorical":[230,240],"ones":[231],"more":[233],"important":[234],"than":[235],"maintaining":[236],"factorized":[237],"variables.":[241]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-03T00:00:00"}