{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:TFYXL6Q2CXOOOAWCUZEJLOKBG6","short_pith_number":"pith:TFYXL6Q2","schema_version":"1.0","canonical_sha256":"997175fa1a15dce702c2a64895b941379947b9007e83a44fb0067b34d46ac7f1","source":{"kind":"arxiv","id":"2203.16262","version":1},"attestation_state":"computed","paper":{"title":"How Does SimSiam Avoid Collapse Without Negative Samples? A Unified Understanding with Self-supervised Contrastive Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chang D. Yoo, Chaoning Zhang, Chenshuang Zhang, In So Kweon, Kang Zhang, Trung X. Pham","submitted_at":"2022-03-30T12:46:31Z","abstract_excerpt":"To avoid collapse in self-supervised learning (SSL), a contrastive loss is widely used but often requires a large number of negative samples. Without negative samples yet achieving competitive performance, a recent work has attracted significant attention for providing a minimalist simple Siamese (SimSiam) method to avoid collapse. However, the reason for how it avoids collapse without negative samples remains not fully clear and our investigation starts by revisiting the explanatory claims in the original SimSiam. After refuting their claims, we introduce vector decomposition for analyzing th"},"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":"2203.16262","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-03-30T12:46:31Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1a7bab93f06556e8232f1a97d8002be9fa5b99ece7c501de6eabb12a4db0995b","abstract_canon_sha256":"7dcacb73f2bee3ab063b35a970183bf1babf8b033b2316e33c1e56dc0b58f0e1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:10:09.299186Z","signature_b64":"iVWDaIvAtpOuX5t2AOL9flOpayd5F1FVcS9EPw47v1H5UiPKNdBC36DoDy8ITjswQCtkishfgR8m1fJbU7UpDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"997175fa1a15dce702c2a64895b941379947b9007e83a44fb0067b34d46ac7f1","last_reissued_at":"2026-07-05T04:10:09.298725Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:10:09.298725Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Does SimSiam Avoid Collapse Without Negative Samples? A Unified Understanding with Self-supervised Contrastive Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chang D. Yoo, Chaoning Zhang, Chenshuang Zhang, In So Kweon, Kang Zhang, Trung X. Pham","submitted_at":"2022-03-30T12:46:31Z","abstract_excerpt":"To avoid collapse in self-supervised learning (SSL), a contrastive loss is widely used but often requires a large number of negative samples. Without negative samples yet achieving competitive performance, a recent work has attracted significant attention for providing a minimalist simple Siamese (SimSiam) method to avoid collapse. However, the reason for how it avoids collapse without negative samples remains not fully clear and our investigation starts by revisiting the explanatory claims in the original SimSiam. After refuting their claims, we introduce vector decomposition for analyzing th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.16262","kind":"arxiv","version":1},"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/2203.16262/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":"2203.16262","created_at":"2026-07-05T04:10:09.298777+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.16262v1","created_at":"2026-07-05T04:10:09.298777+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.16262","created_at":"2026-07-05T04:10:09.298777+00:00"},{"alias_kind":"pith_short_12","alias_value":"TFYXL6Q2CXOO","created_at":"2026-07-05T04:10:09.298777+00:00"},{"alias_kind":"pith_short_16","alias_value":"TFYXL6Q2CXOOOAWC","created_at":"2026-07-05T04:10:09.298777+00:00"},{"alias_kind":"pith_short_8","alias_value":"TFYXL6Q2","created_at":"2026-07-05T04:10:09.298777+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.01443","citing_title":"UR-JEPA: Uniform Rectifiability as a Regularizer for Joint-Embedding Predictive Architectures","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24393","citing_title":"Complexity of Linear Regions in Self-supervised Deep ReLU Networks","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6","json":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6.json","graph_json":"https://pith.science/api/pith-number/TFYXL6Q2CXOOOAWCUZEJLOKBG6/graph.json","events_json":"https://pith.science/api/pith-number/TFYXL6Q2CXOOOAWCUZEJLOKBG6/events.json","paper":"https://pith.science/paper/TFYXL6Q2"},"agent_actions":{"view_html":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6","download_json":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6.json","view_paper":"https://pith.science/paper/TFYXL6Q2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.16262&json=true","fetch_graph":"https://pith.science/api/pith-number/TFYXL6Q2CXOOOAWCUZEJLOKBG6/graph.json","fetch_events":"https://pith.science/api/pith-number/TFYXL6Q2CXOOOAWCUZEJLOKBG6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6/action/storage_attestation","attest_author":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6/action/author_attestation","sign_citation":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6/action/citation_signature","submit_replication":"https://pith.science/pith/TFYXL6Q2CXOOOAWCUZEJLOKBG6/action/replication_record"}},"created_at":"2026-07-05T04:10:09.298777+00:00","updated_at":"2026-07-05T04:10:09.298777+00:00"}