{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:3MR6C5QBNXVAXUCFPWVHR6X632","short_pith_number":"pith:3MR6C5QB","schema_version":"1.0","canonical_sha256":"db23e176016dea0bd0457daa78fafedea988744265dac293cdb963c8d8823b33","source":{"kind":"arxiv","id":"2203.16102","version":1},"attestation_state":"computed","paper":{"title":"Continual Normalization: Rethinking Batch Normalization for Online Continual Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chenghao Liu, Quang Pham, Steven Hoi","submitted_at":"2022-03-30T07:23:24Z","abstract_excerpt":"Existing continual learning methods use Batch Normalization (BN) to facilitate training and improve generalization across tasks. However, the non-i.i.d and non-stationary nature of continual learning data, especially in the online setting, amplify the discrepancy between training and testing in BN and hinder the performance of older tasks. In this work, we study the cross-task normalization effect of BN in online continual learning where BN normalizes the testing data using moments biased towards the current task, resulting in higher catastrophic forgetting. This limitation motivates us to pro"},"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":"2203.16102","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-03-30T07:23:24Z","cross_cats_sorted":[],"title_canon_sha256":"b5ef0909dc2d64343306cc2478520093cb99a357905b715f58c097d32166bc36","abstract_canon_sha256":"f0303ffc86793fa030693b3337306c12d9d338d3462280bfb1a544e8023f3e5d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:10:07.443214Z","signature_b64":"UbBwYPmcrytsq3jdukTowDVjBCrpiPMEIMVzNHM0b2PwoEKo+DBY2mL2Iir/v/6TbhkQtVwRfe9ivs2usMAoBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db23e176016dea0bd0457daa78fafedea988744265dac293cdb963c8d8823b33","last_reissued_at":"2026-07-05T04:10:07.442798Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:10:07.442798Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Continual Normalization: Rethinking Batch Normalization for Online Continual Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chenghao Liu, Quang Pham, Steven Hoi","submitted_at":"2022-03-30T07:23:24Z","abstract_excerpt":"Existing continual learning methods use Batch Normalization (BN) to facilitate training and improve generalization across tasks. However, the non-i.i.d and non-stationary nature of continual learning data, especially in the online setting, amplify the discrepancy between training and testing in BN and hinder the performance of older tasks. In this work, we study the cross-task normalization effect of BN in online continual learning where BN normalizes the testing data using moments biased towards the current task, resulting in higher catastrophic forgetting. This limitation motivates us to pro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.16102","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/2203.16102/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":"2203.16102","created_at":"2026-07-05T04:10:07.442863+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.16102v1","created_at":"2026-07-05T04:10:07.442863+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.16102","created_at":"2026-07-05T04:10:07.442863+00:00"},{"alias_kind":"pith_short_12","alias_value":"3MR6C5QBNXVA","created_at":"2026-07-05T04:10:07.442863+00:00"},{"alias_kind":"pith_short_16","alias_value":"3MR6C5QBNXVAXUCF","created_at":"2026-07-05T04:10:07.442863+00:00"},{"alias_kind":"pith_short_8","alias_value":"3MR6C5QB","created_at":"2026-07-05T04:10:07.442863+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.27482","citing_title":"Energy-Structured Low-Rank Adaptation for Continual Learning","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22562","citing_title":"Activation Function Design Sustains Plasticity in Continual Learning","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632","json":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632.json","graph_json":"https://pith.science/api/pith-number/3MR6C5QBNXVAXUCFPWVHR6X632/graph.json","events_json":"https://pith.science/api/pith-number/3MR6C5QBNXVAXUCFPWVHR6X632/events.json","paper":"https://pith.science/paper/3MR6C5QB"},"agent_actions":{"view_html":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632","download_json":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632.json","view_paper":"https://pith.science/paper/3MR6C5QB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.16102&json=true","fetch_graph":"https://pith.science/api/pith-number/3MR6C5QBNXVAXUCFPWVHR6X632/graph.json","fetch_events":"https://pith.science/api/pith-number/3MR6C5QBNXVAXUCFPWVHR6X632/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632/action/storage_attestation","attest_author":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632/action/author_attestation","sign_citation":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632/action/citation_signature","submit_replication":"https://pith.science/pith/3MR6C5QBNXVAXUCFPWVHR6X632/action/replication_record"}},"created_at":"2026-07-05T04:10:07.442863+00:00","updated_at":"2026-07-05T04:10:07.442863+00:00"}