{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:6B6RKBZPUNOIJM6QPG6NETQ3EP","short_pith_number":"pith:6B6RKBZP","schema_version":"1.0","canonical_sha256":"f07d15072fa35c84b3d079bcd24e1b23e08a000d3139a35b2e9c5414e21aa3b0","source":{"kind":"arxiv","id":"2111.00743","version":4},"attestation_state":"computed","paper":{"title":"Towards the Generalization of Contrastive Self-Supervised Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Mingyang Yi, Weiran Huang, Xuyang Zhao, Zihao Jiang","submitted_at":"2021-11-01T07:39:38Z","abstract_excerpt":"Recently, self-supervised learning has attracted great attention, since it only requires unlabeled data for model training. Contrastive learning is one popular method for self-supervised learning and has achieved promising empirical performance. However, the theoretical understanding of its generalization ability is still limited. To this end, we define a kind of $(\\sigma,\\delta)$-measure to mathematically quantify the data augmentation, and then provide an upper bound of the downstream classification error rate based on the measure. It reveals that the generalization ability of contrastive se"},"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":"2111.00743","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-11-01T07:39:38Z","cross_cats_sorted":["cs.AI","cs.CV","stat.ML"],"title_canon_sha256":"f8d12b31680cc4526051d1ca9664c1fdcdeb93fd74f8079cf040e762596d1a1b","abstract_canon_sha256":"2a536c4a38739f6467932de6c87be7e67e08412346daee2995ef9210cfba7aab"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:47:14.350542Z","signature_b64":"bXwqxJ/rmz0eE9qpU1m+MroCyJm2naxBUZlNGfLo2WLDSDPwQMxr5XLrcSY09vr1KXiThJEL2SjtglzF56g3Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f07d15072fa35c84b3d079bcd24e1b23e08a000d3139a35b2e9c5414e21aa3b0","last_reissued_at":"2026-07-05T05:47:14.350086Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:47:14.350086Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards the Generalization of Contrastive Self-Supervised Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Mingyang Yi, Weiran Huang, Xuyang Zhao, Zihao Jiang","submitted_at":"2021-11-01T07:39:38Z","abstract_excerpt":"Recently, self-supervised learning has attracted great attention, since it only requires unlabeled data for model training. Contrastive learning is one popular method for self-supervised learning and has achieved promising empirical performance. However, the theoretical understanding of its generalization ability is still limited. To this end, we define a kind of $(\\sigma,\\delta)$-measure to mathematically quantify the data augmentation, and then provide an upper bound of the downstream classification error rate based on the measure. It reveals that the generalization ability of contrastive se"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.00743","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/2111.00743/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":"2111.00743","created_at":"2026-07-05T05:47:14.350144+00:00"},{"alias_kind":"arxiv_version","alias_value":"2111.00743v4","created_at":"2026-07-05T05:47:14.350144+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.00743","created_at":"2026-07-05T05:47:14.350144+00:00"},{"alias_kind":"pith_short_12","alias_value":"6B6RKBZPUNOI","created_at":"2026-07-05T05:47:14.350144+00:00"},{"alias_kind":"pith_short_16","alias_value":"6B6RKBZPUNOIJM6Q","created_at":"2026-07-05T05:47:14.350144+00:00"},{"alias_kind":"pith_short_8","alias_value":"6B6RKBZP","created_at":"2026-07-05T05:47:14.350144+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22873","citing_title":"SingGuard: A Policy-Adaptive Multimodal LLM Guardrail with Dynamic Reasoning","ref_index":114,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22873","citing_title":"SingGuard: A Policy-Adaptive Multimodal LLM Guardrail with Dynamic Reasoning","ref_index":114,"is_internal_anchor":false},{"citing_arxiv_id":"2606.18333","citing_title":"Polarisation and Faraday rotation measure imaging at metre wavelengths with sub-arcsecond resolution: a foundational calibration strategy","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2502.05075","citing_title":"Discrepancies are Virtue: Weak-to-Strong Generalization through Lens of Intrinsic Dimension","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2602.06382","citing_title":"Now You See That: Learning End-to-End Humanoid Locomotion from Raw Pixels","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP","json":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP.json","graph_json":"https://pith.science/api/pith-number/6B6RKBZPUNOIJM6QPG6NETQ3EP/graph.json","events_json":"https://pith.science/api/pith-number/6B6RKBZPUNOIJM6QPG6NETQ3EP/events.json","paper":"https://pith.science/paper/6B6RKBZP"},"agent_actions":{"view_html":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP","download_json":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP.json","view_paper":"https://pith.science/paper/6B6RKBZP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2111.00743&json=true","fetch_graph":"https://pith.science/api/pith-number/6B6RKBZPUNOIJM6QPG6NETQ3EP/graph.json","fetch_events":"https://pith.science/api/pith-number/6B6RKBZPUNOIJM6QPG6NETQ3EP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP/action/storage_attestation","attest_author":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP/action/author_attestation","sign_citation":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP/action/citation_signature","submit_replication":"https://pith.science/pith/6B6RKBZPUNOIJM6QPG6NETQ3EP/action/replication_record"}},"created_at":"2026-07-05T05:47:14.350144+00:00","updated_at":"2026-07-05T05:47:14.350144+00:00"}