{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:CAU3CFENXU42YQBUHSQRXUHG4L","short_pith_number":"pith:CAU3CFEN","schema_version":"1.0","canonical_sha256":"1029b1148dbd39ac40343ca11bd0e6e2d3ca0d8fcead8e4d52ae4f7c4ae20189","source":{"kind":"arxiv","id":"2402.09542","version":3},"attestation_state":"computed","paper":{"title":"Layerwise Proximal Replay: A Proximal Point Method for Online Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Frank Wood, Geoff Pleiss, Jason Yoo, Yunpeng Liu","submitted_at":"2024-02-14T19:34:28Z","abstract_excerpt":"In online continual learning, a neural network incrementally learns from a non-i.i.d. data stream. Nearly all online continual learning methods employ experience replay to simultaneously prevent catastrophic forgetting and underfitting on past data. Our work demonstrates a limitation of this approach: neural networks trained with experience replay tend to have unstable optimization trajectories, impeding their overall accuracy. Surprisingly, these instabilities persist even when the replay buffer stores all previous training examples, suggesting that this issue is orthogonal to catastrophic fo"},"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":"2402.09542","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-02-14T19:34:28Z","cross_cats_sorted":[],"title_canon_sha256":"07262a10cff22c7c36c574797f015f6a60aeff0525915f3bcfc6624427087b7f","abstract_canon_sha256":"6dd42157206c6ba073bb287135253aeca058f437a5ad87b6f5f83016ccc4d868"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:45:49.989339Z","signature_b64":"RYhsaCwFtOAxGj2VbPvj0rMKTa64VG2OhEeCMvxRwcH3iPz0DvUR2QH9hQxDcQQmTVZEaF+VDlEOOHZUZu4ODA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1029b1148dbd39ac40343ca11bd0e6e2d3ca0d8fcead8e4d52ae4f7c4ae20189","last_reissued_at":"2026-07-05T08:45:49.988900Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:45:49.988900Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Layerwise Proximal Replay: A Proximal Point Method for Online Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Frank Wood, Geoff Pleiss, Jason Yoo, Yunpeng Liu","submitted_at":"2024-02-14T19:34:28Z","abstract_excerpt":"In online continual learning, a neural network incrementally learns from a non-i.i.d. data stream. Nearly all online continual learning methods employ experience replay to simultaneously prevent catastrophic forgetting and underfitting on past data. Our work demonstrates a limitation of this approach: neural networks trained with experience replay tend to have unstable optimization trajectories, impeding their overall accuracy. Surprisingly, these instabilities persist even when the replay buffer stores all previous training examples, suggesting that this issue is orthogonal to catastrophic fo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.09542","kind":"arxiv","version":3},"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/2402.09542/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":"2402.09542","created_at":"2026-07-05T08:45:49.988955+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.09542v3","created_at":"2026-07-05T08:45:49.988955+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.09542","created_at":"2026-07-05T08:45:49.988955+00:00"},{"alias_kind":"pith_short_12","alias_value":"CAU3CFENXU42","created_at":"2026-07-05T08:45:49.988955+00:00"},{"alias_kind":"pith_short_16","alias_value":"CAU3CFENXU42YQBU","created_at":"2026-07-05T08:45:49.988955+00:00"},{"alias_kind":"pith_short_8","alias_value":"CAU3CFEN","created_at":"2026-07-05T08:45:49.988955+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.01866","citing_title":"Online Curvature-Aware Replay: Leveraging $\\mathbf{2^{nd}}$ Order Information for Online Continual Learning","ref_index":60,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L","json":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L.json","graph_json":"https://pith.science/api/pith-number/CAU3CFENXU42YQBUHSQRXUHG4L/graph.json","events_json":"https://pith.science/api/pith-number/CAU3CFENXU42YQBUHSQRXUHG4L/events.json","paper":"https://pith.science/paper/CAU3CFEN"},"agent_actions":{"view_html":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L","download_json":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L.json","view_paper":"https://pith.science/paper/CAU3CFEN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.09542&json=true","fetch_graph":"https://pith.science/api/pith-number/CAU3CFENXU42YQBUHSQRXUHG4L/graph.json","fetch_events":"https://pith.science/api/pith-number/CAU3CFENXU42YQBUHSQRXUHG4L/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L/action/storage_attestation","attest_author":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L/action/author_attestation","sign_citation":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L/action/citation_signature","submit_replication":"https://pith.science/pith/CAU3CFENXU42YQBUHSQRXUHG4L/action/replication_record"}},"created_at":"2026-07-05T08:45:49.988955+00:00","updated_at":"2026-07-05T08:45:49.988955+00:00"}