{"id":"https://openalex.org/W7163152452","doi":"https://doi.org/10.48550/arxiv.2606.02572","title":"VISReg: Variance-Invariance-Sketching Regularization for JEPA training","display_name":"VISReg: Variance-Invariance-Sketching Regularization for JEPA training","publication_year":2026,"publication_date":"2026-06-01","ids":{"openalex":"https://openalex.org/W7163152452","doi":"https://doi.org/10.48550/arxiv.2606.02572"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.02572","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.02572","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":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.02572","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5026177673","display_name":"Haiyu Wu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wu, Haiyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137671135","display_name":"Randall Balestriero","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Balestriero, Randall","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137622195","display_name":"Morgan Levine","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Levine, Morgan","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/T10036","display_name":"Advanced Neural Network Applications","score":0.4068000018596649,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.4068000018596649,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.210099995136261,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.1420000046491623,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/regularization","display_name":"Regularization (linguistics)","score":0.6916000247001648},{"id":"https://openalex.org/keywords/covariance","display_name":"Covariance","score":0.6568999886512756},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.5925999879837036},{"id":"https://openalex.org/keywords/tikhonov-regularization","display_name":"Tikhonov regularization","score":0.5551999807357788},{"id":"https://openalex.org/keywords/heuristics","display_name":"Heuristics","score":0.42149999737739563},{"id":"https://openalex.org/keywords/decoupling","display_name":"Decoupling (probability)","score":0.35569998621940613},{"id":"https://openalex.org/keywords/flexibility","display_name":"Flexibility (engineering)","score":0.34299999475479126},{"id":"https://openalex.org/keywords/variance","display_name":"Variance (accounting)","score":0.33250001072883606}],"concepts":[{"id":"https://openalex.org/C2776135515","wikidata":"https://www.wikidata.org/wiki/Q17143721","display_name":"Regularization (linguistics)","level":2,"score":0.6916000247001648},{"id":"https://openalex.org/C178650346","wikidata":"https://www.wikidata.org/wiki/Q201984","display_name":"Covariance","level":2,"score":0.6568999886512756},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.5925999879837036},{"id":"https://openalex.org/C152442038","wikidata":"https://www.wikidata.org/wiki/Q2778212","display_name":"Tikhonov regularization","level":3,"score":0.5551999807357788},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5450999736785889},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.47859999537467957},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.42149999737739563},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.41830000281333923},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.4138999879360199},{"id":"https://openalex.org/C205606062","wikidata":"https://www.wikidata.org/wiki/Q5249645","display_name":"Decoupling (probability)","level":2,"score":0.35569998621940613},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3549000024795532},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.34299999475479126},{"id":"https://openalex.org/C196083921","wikidata":"https://www.wikidata.org/wiki/Q7915758","display_name":"Variance (accounting)","level":2,"score":0.33250001072883606},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.3248000144958496},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.3239000141620636},{"id":"https://openalex.org/C177860922","wikidata":"https://www.wikidata.org/wiki/Q788608","display_name":"Decorrelation","level":2,"score":0.32100000977516174},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.31150001287460327},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.31130000948905945},{"id":"https://openalex.org/C197640229","wikidata":"https://www.wikidata.org/wiki/Q2534066","display_name":"Predictability","level":2,"score":0.30730000138282776},{"id":"https://openalex.org/C119340705","wikidata":"https://www.wikidata.org/wiki/Q1628597","display_name":"Analysis of covariance","level":2,"score":0.28850001096725464},{"id":"https://openalex.org/C135252773","wikidata":"https://www.wikidata.org/wiki/Q1567213","display_name":"Inverse problem","level":2,"score":0.2777999937534332},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.2639000117778778},{"id":"https://openalex.org/C154881674","wikidata":"https://www.wikidata.org/wiki/Q2269270","display_name":"Variogram","level":3,"score":0.2606000006198883}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.02572","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.02572","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":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.02572","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.02572","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":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Self-supervised":[0],"learning":[1],"methods":[2,53],"prevent":[3],"embedding":[4,14],"collapse":[5],"via":[6],"modeling":[7],"heuristics":[8],"or":[9],"explicit":[10],"regularization":[11,21,132],"of":[12,115],"the":[13,17,44,112,163],"space.":[15],"Among":[16],"latter,":[18],"VICReg":[19],"decomposes":[20],"into":[22],"variance":[23,97],"and":[24,29,69,105,136,141,171],"covariance":[25,32,83],"objectives,":[26],"offering":[27],"flexibility":[28,68,110],"interpretability.":[30],"However,":[31],"captures":[33],"only":[34],"second-order":[35],"statistics":[36],"--":[37],"encouraging":[38],"decorrelation":[39],"but":[40,66],"failing":[41],"to":[42,62,139],"enforce":[43],"full":[45,91],"distributional":[46,92,113],"shape":[47],"needed":[48],"for":[49,99],"stable":[50],"training.":[51],"Sketching-based":[52],"such":[54],"as":[55],"SIGReg":[56],"address":[57],"this":[58],"by":[59],"aligning":[60],"embeddings":[61],"an":[63],"isotropic":[64],"Gaussian,":[65],"lack":[67],"suffer":[70],"from":[71],"vanishing":[72],"gradients":[73,120],"under":[74,122],"collapse.":[75,123],"We":[76,124],"propose":[77],"Variance-Invariance-Sketching":[78],"Regularization":[79],"(VISReg),":[80],"which":[81],"replaces":[82],"with":[84,111],"a":[85,96],"Sliced-Wasserstein-based":[86],"sketching":[87,116],"objective":[88],"that":[89,126],"enforces":[90],"shape,":[93,106],"while":[94],"retaining":[95],"term":[98],"scale":[100,104],"control.":[101],"By":[102],"decoupling":[103],"VISReg":[107,127,147],"combines":[108],"VICReg's":[109],"rigor":[114],"methods,":[117],"providing":[118],"robust":[119],"even":[121],"show":[125],"scales":[128],"linearly,":[129],"outperforms":[130],"existing":[131],"on":[133,145,151,155],"low-quality":[134],"datasets,":[135],"is":[137],"resilient":[138],"long-tailed":[140],"low-rank":[142],"regimes.":[143],"Pre-trained":[144,154],"ImageNet-1K,":[146],"achieves":[148],"state-of-the-art":[149],"performance":[150,161],"out-of-distribution":[152],"datasets.":[153],"ImageNet-22K,":[156],"it":[157],"matches":[158],"DINOv2's":[159],"OOD":[160],"despite":[162],"latter":[164],"using":[165],"10x":[166],"more":[167],"data":[168],"(LVD-142M).":[169],"Project":[170],"code:":[172],"https://haiyuwu.github.io/visreg.":[173]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-03T00:00:00"}