{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:OFMG3K653PXMVIB5RDG76I7OKO","short_pith_number":"pith:OFMG3K65","schema_version":"1.0","canonical_sha256":"71586dabdddbeecaa03d88cdff23ee53bd5c901235b50eeef78e7c9ac76468ff","source":{"kind":"arxiv","id":"2210.01883","version":2},"attestation_state":"computed","paper":{"title":"Contrastive Learning Can Find An Optimal Basis For Approximately View-Invariant Functions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ayoub El Hanchi, Chris J. Maddison, Daniel D. Johnson","submitted_at":"2022-10-04T20:02:52Z","abstract_excerpt":"Contrastive learning is a powerful framework for learning self-supervised representations that generalize well to downstream supervised tasks. We show that multiple existing contrastive learning methods can be reinterpreted as learning kernel functions that approximate a fixed positive-pair kernel. We then prove that a simple representation obtained by combining this kernel with PCA provably minimizes the worst-case approximation error of linear predictors, under a straightforward assumption that positive pairs have similar labels. Our analysis is based on a decomposition of the target functio"},"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":"2210.01883","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-10-04T20:02:52Z","cross_cats_sorted":[],"title_canon_sha256":"734335d9f9ebd1f57858cdbc5ea1519d65d9d0d8a2401ccfe380de9c64650d5c","abstract_canon_sha256":"2a145bd866986f859aa6b8a0268e13d374956af90ea42baba05c5bae396d7ba1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:41:51.198297Z","signature_b64":"WYBzikmY5BFayYshImvSo9MY34WHsiWvHpDQ4wOgu5D6Iku3AWwLF04mCx9lAuVotc516PwTskat7UeEfjrNDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"71586dabdddbeecaa03d88cdff23ee53bd5c901235b50eeef78e7c9ac76468ff","last_reissued_at":"2026-07-05T05:41:51.197853Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:41:51.197853Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Contrastive Learning Can Find An Optimal Basis For Approximately View-Invariant Functions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ayoub El Hanchi, Chris J. Maddison, Daniel D. Johnson","submitted_at":"2022-10-04T20:02:52Z","abstract_excerpt":"Contrastive learning is a powerful framework for learning self-supervised representations that generalize well to downstream supervised tasks. We show that multiple existing contrastive learning methods can be reinterpreted as learning kernel functions that approximate a fixed positive-pair kernel. We then prove that a simple representation obtained by combining this kernel with PCA provably minimizes the worst-case approximation error of linear predictors, under a straightforward assumption that positive pairs have similar labels. Our analysis is based on a decomposition of the target functio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.01883","kind":"arxiv","version":2},"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/2210.01883/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":"2210.01883","created_at":"2026-07-05T05:41:51.197910+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.01883v2","created_at":"2026-07-05T05:41:51.197910+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.01883","created_at":"2026-07-05T05:41:51.197910+00:00"},{"alias_kind":"pith_short_12","alias_value":"OFMG3K653PXM","created_at":"2026-07-05T05:41:51.197910+00:00"},{"alias_kind":"pith_short_16","alias_value":"OFMG3K653PXMVIB5","created_at":"2026-07-05T05:41:51.197910+00:00"},{"alias_kind":"pith_short_8","alias_value":"OFMG3K65","created_at":"2026-07-05T05:41:51.197910+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21590","citing_title":"Radial Basis Function Networks as Projection Heads in Self-Supervised Learning","ref_index":27,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO","json":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO.json","graph_json":"https://pith.science/api/pith-number/OFMG3K653PXMVIB5RDG76I7OKO/graph.json","events_json":"https://pith.science/api/pith-number/OFMG3K653PXMVIB5RDG76I7OKO/events.json","paper":"https://pith.science/paper/OFMG3K65"},"agent_actions":{"view_html":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO","download_json":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO.json","view_paper":"https://pith.science/paper/OFMG3K65","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.01883&json=true","fetch_graph":"https://pith.science/api/pith-number/OFMG3K653PXMVIB5RDG76I7OKO/graph.json","fetch_events":"https://pith.science/api/pith-number/OFMG3K653PXMVIB5RDG76I7OKO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO/action/storage_attestation","attest_author":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO/action/author_attestation","sign_citation":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO/action/citation_signature","submit_replication":"https://pith.science/pith/OFMG3K653PXMVIB5RDG76I7OKO/action/replication_record"}},"created_at":"2026-07-05T05:41:51.197910+00:00","updated_at":"2026-07-05T05:41:51.197910+00:00"}