{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:UBM34LZ7Q7J5WMXRCZPSSLNAYI","short_pith_number":"pith:UBM34LZ7","schema_version":"1.0","canonical_sha256":"a059be2f3f87d3db32f1165f292da0c2293efaabd01daf38ce1ed969e2a40c22","source":{"kind":"arxiv","id":"2203.14544","version":1},"attestation_state":"computed","paper":{"title":"Gradient-Matching Coresets for Rehearsal-Based Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"C\\'edric Archambeau, Giovanni Zappella, Lukas Balles","submitted_at":"2022-03-28T07:37:17Z","abstract_excerpt":"The goal of continual learning (CL) is to efficiently update a machine learning model with new data without forgetting previously-learned knowledge. Most widely-used CL methods rely on a rehearsal memory of data points to be reused while training on new data. Curating such a rehearsal memory to maintain a small, informative subset of all the data seen so far is crucial to the success of these methods. We devise a coreset selection method for rehearsal-based continual learning. Our method is based on the idea of gradient matching: The gradients induced by the coreset should match, as closely as"},"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":"2203.14544","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-03-28T07:37:17Z","cross_cats_sorted":[],"title_canon_sha256":"eb3aabf5e61c927db5ff5353bfbe026e94f7ec4254d435099f88395b90763888","abstract_canon_sha256":"d3cbb9e6ba78b7c1dd49fab5c1ef4bc469fa2efa8b9f52d2a681026cf254be26"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:09:06.320238Z","signature_b64":"j+it+1k05Oi385IKqgE2MrvVcU1gChZwHtDXrE49Ph58Qq5SMnnXVx/74YEnzL/Ihc7kiV/mqhSrSYNbHyrUCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a059be2f3f87d3db32f1165f292da0c2293efaabd01daf38ce1ed969e2a40c22","last_reissued_at":"2026-07-05T04:09:06.319824Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:09:06.319824Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradient-Matching Coresets for Rehearsal-Based Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"C\\'edric Archambeau, Giovanni Zappella, Lukas Balles","submitted_at":"2022-03-28T07:37:17Z","abstract_excerpt":"The goal of continual learning (CL) is to efficiently update a machine learning model with new data without forgetting previously-learned knowledge. Most widely-used CL methods rely on a rehearsal memory of data points to be reused while training on new data. Curating such a rehearsal memory to maintain a small, informative subset of all the data seen so far is crucial to the success of these methods. We devise a coreset selection method for rehearsal-based continual learning. Our method is based on the idea of gradient matching: The gradients induced by the coreset should match, as closely as"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.14544","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/2203.14544/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":"2203.14544","created_at":"2026-07-05T04:09:06.319891+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.14544v1","created_at":"2026-07-05T04:09:06.319891+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.14544","created_at":"2026-07-05T04:09:06.319891+00:00"},{"alias_kind":"pith_short_12","alias_value":"UBM34LZ7Q7J5","created_at":"2026-07-05T04:09:06.319891+00:00"},{"alias_kind":"pith_short_16","alias_value":"UBM34LZ7Q7J5WMXR","created_at":"2026-07-05T04:09:06.319891+00:00"},{"alias_kind":"pith_short_8","alias_value":"UBM34LZ7","created_at":"2026-07-05T04:09:06.319891+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.09418","citing_title":"Domain-Generalization to Improve Learning in Meta-Learning Algorithms","ref_index":47,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI","json":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI.json","graph_json":"https://pith.science/api/pith-number/UBM34LZ7Q7J5WMXRCZPSSLNAYI/graph.json","events_json":"https://pith.science/api/pith-number/UBM34LZ7Q7J5WMXRCZPSSLNAYI/events.json","paper":"https://pith.science/paper/UBM34LZ7"},"agent_actions":{"view_html":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI","download_json":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI.json","view_paper":"https://pith.science/paper/UBM34LZ7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.14544&json=true","fetch_graph":"https://pith.science/api/pith-number/UBM34LZ7Q7J5WMXRCZPSSLNAYI/graph.json","fetch_events":"https://pith.science/api/pith-number/UBM34LZ7Q7J5WMXRCZPSSLNAYI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI/action/storage_attestation","attest_author":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI/action/author_attestation","sign_citation":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI/action/citation_signature","submit_replication":"https://pith.science/pith/UBM34LZ7Q7J5WMXRCZPSSLNAYI/action/replication_record"}},"created_at":"2026-07-05T04:09:06.319891+00:00","updated_at":"2026-07-05T04:09:06.319891+00:00"}