{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:VOUYRAJZE7IACBT4YS5P5BZ36H","short_pith_number":"pith:VOUYRAJZ","schema_version":"1.0","canonical_sha256":"aba988813927d001067cc4bafe873bf1c8df377e7cfed42de1d43c9b97c32e88","source":{"kind":"arxiv","id":"2106.04619","version":4},"attestation_state":"computed","paper":{"title":"Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.LG"],"primary_cat":"stat.ML","authors_text":"Bernhard Sch\\\"olkopf, Francesco Locatello, Julius von K\\\"ugelgen, Luigi Gresele, Michel Besserve, Wieland Brendel, Yash Sharma","submitted_at":"2021-06-08T18:18:09Z","abstract_excerpt":"Self-supervised representation learning has shown remarkable success in a number of domains. A common practice is to perform data augmentation via hand-crafted transformations intended to leave the semantics of the data invariant. We seek to understand the empirical success of this approach from a theoretical perspective. We formulate the augmentation process as a latent variable model by postulating a partition of the latent representation into a content component, which is assumed invariant to augmentation, and a style component, which is allowed to change. Unlike prior work on disentangleme"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2106.04619","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2021-06-08T18:18:09Z","cross_cats_sorted":["cs.AI","cs.CV","cs.LG"],"title_canon_sha256":"7b65ec1d2c3f14f65a8fea8f056e76f9db89ffe3d0ea0bd9c091f708a11be5e0","abstract_canon_sha256":"efc72f1354486a70a5c47e1108c6ab3ed9e87a39cdbd2e9fce683d04522db8d3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:48:29.622819Z","signature_b64":"OIk2lreBu+ANo61weMO37ktFuxWqyIW2KIYSxo441dYJAAtWYAWPfe6ZNGdQE0m6HmOEOCTuDWzULGBeo/S2Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aba988813927d001067cc4bafe873bf1c8df377e7cfed42de1d43c9b97c32e88","last_reissued_at":"2026-07-05T03:48:29.622370Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:48:29.622370Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-Supervised Learning with Data Augmentations Provably Isolates Content from Style","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.LG"],"primary_cat":"stat.ML","authors_text":"Bernhard Sch\\\"olkopf, Francesco Locatello, Julius von K\\\"ugelgen, Luigi Gresele, Michel Besserve, Wieland Brendel, Yash Sharma","submitted_at":"2021-06-08T18:18:09Z","abstract_excerpt":"Self-supervised representation learning has shown remarkable success in a number of domains. A common practice is to perform data augmentation via hand-crafted transformations intended to leave the semantics of the data invariant. We seek to understand the empirical success of this approach from a theoretical perspective. We formulate the augmentation process as a latent variable model by postulating a partition of the latent representation into a content component, which is assumed invariant to augmentation, and a style component, which is allowed to change. Unlike prior work on disentangleme"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.04619","kind":"arxiv","version":4},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2106.04619/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2106.04619","created_at":"2026-07-05T03:48:29.622434+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.04619v4","created_at":"2026-07-05T03:48:29.622434+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.04619","created_at":"2026-07-05T03:48:29.622434+00:00"},{"alias_kind":"pith_short_12","alias_value":"VOUYRAJZE7IA","created_at":"2026-07-05T03:48:29.622434+00:00"},{"alias_kind":"pith_short_16","alias_value":"VOUYRAJZE7IACBT4","created_at":"2026-07-05T03:48:29.622434+00:00"},{"alias_kind":"pith_short_8","alias_value":"VOUYRAJZ","created_at":"2026-07-05T03:48:29.622434+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.12733","citing_title":"From Generalist to Specialist Representation","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17568","citing_title":"Diverse Dictionary Learning","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H","json":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H.json","graph_json":"https://pith.science/api/pith-number/VOUYRAJZE7IACBT4YS5P5BZ36H/graph.json","events_json":"https://pith.science/api/pith-number/VOUYRAJZE7IACBT4YS5P5BZ36H/events.json","paper":"https://pith.science/paper/VOUYRAJZ"},"agent_actions":{"view_html":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H","download_json":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H.json","view_paper":"https://pith.science/paper/VOUYRAJZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.04619&json=true","fetch_graph":"https://pith.science/api/pith-number/VOUYRAJZE7IACBT4YS5P5BZ36H/graph.json","fetch_events":"https://pith.science/api/pith-number/VOUYRAJZE7IACBT4YS5P5BZ36H/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H/action/storage_attestation","attest_author":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H/action/author_attestation","sign_citation":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H/action/citation_signature","submit_replication":"https://pith.science/pith/VOUYRAJZE7IACBT4YS5P5BZ36H/action/replication_record"}},"created_at":"2026-07-05T03:48:29.622434+00:00","updated_at":"2026-07-05T03:48:29.622434+00:00"}