{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DDHSJGMQZK2BI5XQRLSWNJNH7I","short_pith_number":"pith:DDHSJGMQ","schema_version":"1.0","canonical_sha256":"18cf249990cab41476f08ae566a5a7fa3b99bd82851c9950a0e0795822c59d04","source":{"kind":"arxiv","id":"2502.01866","version":1},"attestation_state":"computed","paper":{"title":"Online Curvature-Aware Replay: Leveraging $\\mathbf{2^{nd}}$ Order Information for Online Continual Learning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Antonio Carta, Edoardo Urettini","submitted_at":"2025-02-03T22:31:36Z","abstract_excerpt":"Online Continual Learning (OCL) models continuously adapt to nonstationary data streams, usually without task information. These settings are complex and many traditional CL methods fail, while online methods (mainly replay-based) suffer from instabilities after the task shift. To address this issue, we formalize replay-based OCL as a second-order online joint optimization with explicit KL-divergence constraints on replay data. We propose Online Curvature-Aware Replay (OCAR) to solve the problem: a method that leverages second-order information of the loss using a K-FAC approximation of the Fi"},"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":"2502.01866","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-03T22:31:36Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"cdaf323a6a25a85e861e3fbc204e96c4e3d3054c5e08ba01e9854e164795667d","abstract_canon_sha256":"77a442bf77f6ec670d501bac0809d0b2e97ab199da9e45fc404be90263902bbe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:09:20.460359Z","signature_b64":"ElZCUjfCiOVvu0kYPyR5/GOJIS27tvpyPn/9bD2/aYRTDi0/A79WZ9rpqWhNQFipenis1zluEvuewCHtE4owAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"18cf249990cab41476f08ae566a5a7fa3b99bd82851c9950a0e0795822c59d04","last_reissued_at":"2026-07-05T10:09:20.459982Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:09:20.459982Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Online Curvature-Aware Replay: Leveraging $\\mathbf{2^{nd}}$ Order Information for Online Continual Learning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Antonio Carta, Edoardo Urettini","submitted_at":"2025-02-03T22:31:36Z","abstract_excerpt":"Online Continual Learning (OCL) models continuously adapt to nonstationary data streams, usually without task information. These settings are complex and many traditional CL methods fail, while online methods (mainly replay-based) suffer from instabilities after the task shift. To address this issue, we formalize replay-based OCL as a second-order online joint optimization with explicit KL-divergence constraints on replay data. We propose Online Curvature-Aware Replay (OCAR) to solve the problem: a method that leverages second-order information of the loss using a K-FAC approximation of the Fi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.01866","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/2502.01866/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":"2502.01866","created_at":"2026-07-05T10:09:20.460036+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.01866v1","created_at":"2026-07-05T10:09:20.460036+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.01866","created_at":"2026-07-05T10:09:20.460036+00:00"},{"alias_kind":"pith_short_12","alias_value":"DDHSJGMQZK2B","created_at":"2026-07-05T10:09:20.460036+00:00"},{"alias_kind":"pith_short_16","alias_value":"DDHSJGMQZK2BI5XQ","created_at":"2026-07-05T10:09:20.460036+00:00"},{"alias_kind":"pith_short_8","alias_value":"DDHSJGMQ","created_at":"2026-07-05T10:09:20.460036+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I","json":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I.json","graph_json":"https://pith.science/api/pith-number/DDHSJGMQZK2BI5XQRLSWNJNH7I/graph.json","events_json":"https://pith.science/api/pith-number/DDHSJGMQZK2BI5XQRLSWNJNH7I/events.json","paper":"https://pith.science/paper/DDHSJGMQ"},"agent_actions":{"view_html":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I","download_json":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I.json","view_paper":"https://pith.science/paper/DDHSJGMQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.01866&json=true","fetch_graph":"https://pith.science/api/pith-number/DDHSJGMQZK2BI5XQRLSWNJNH7I/graph.json","fetch_events":"https://pith.science/api/pith-number/DDHSJGMQZK2BI5XQRLSWNJNH7I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I/action/storage_attestation","attest_author":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I/action/author_attestation","sign_citation":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I/action/citation_signature","submit_replication":"https://pith.science/pith/DDHSJGMQZK2BI5XQRLSWNJNH7I/action/replication_record"}},"created_at":"2026-07-05T10:09:20.460036+00:00","updated_at":"2026-07-05T10:09:20.460036+00:00"}