{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YKRUGGKT2JX63MZNVZH7ODLEKB","short_pith_number":"pith:YKRUGGKT","schema_version":"1.0","canonical_sha256":"c2a3431953d26fedb32dae4ff70d64504c8d501013beeb61e265a785ea869016","source":{"kind":"arxiv","id":"2502.11927","version":1},"attestation_state":"computed","paper":{"title":"Continual Learning Should Move Beyond Incremental Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Antonio Alliegro, Antonio Carta, Carlo D'Eramo, Dustin Carri\\'on-Ojeda, Fabian Hinder, Francesca Pistilli, Georgia Chalvatzaki, Gido M van de Ven, Kristian Kersting, Martin Mundt, Nikhil Churamani, Raffaello Camoriano, Robin Hesse, Roshni Ramanna Kamath, Rupert Mitchell, Samin Hamidi, Simone Schaub-Meyer, Subarnaduti Paul, Tinne Tuytelaars, Vincenzo Lomonaco","submitted_at":"2025-02-17T15:40:13Z","abstract_excerpt":"Continual learning (CL) is the sub-field of machine learning concerned with accumulating knowledge in dynamic environments. So far, CL research has mainly focused on incremental classification tasks, where models learn to classify new categories while retaining knowledge of previously learned ones. Here, we argue that maintaining such a focus limits both theoretical development and practical applicability of CL methods. Through a detailed analysis of concrete examples - including multi-target classification, robotics with constrained output spaces, learning in continuous task domains, and high"},"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":"2502.11927","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-17T15:40:13Z","cross_cats_sorted":[],"title_canon_sha256":"693f8eec9d2197e2c034f5955137b136a158d090fee2b7f110f80873ab1492ef","abstract_canon_sha256":"ac3a9734a17c22673ea8a6f9acbfd3687774f2985e7b0fa63ead1a86a3ae4127"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:36.300100Z","signature_b64":"LsJWuxFLh9n5WJGfS7bBwgS+rnt8SeDJMbi/KF4v9VRH1B9Dh0Uhyf8fHv+s61zrbLSl4nDkhkJwVZIJr6uuBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c2a3431953d26fedb32dae4ff70d64504c8d501013beeb61e265a785ea869016","last_reissued_at":"2026-07-05T10:15:36.299605Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:36.299605Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Continual Learning Should Move Beyond Incremental Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Antonio Alliegro, Antonio Carta, Carlo D'Eramo, Dustin Carri\\'on-Ojeda, Fabian Hinder, Francesca Pistilli, Georgia Chalvatzaki, Gido M van de Ven, Kristian Kersting, Martin Mundt, Nikhil Churamani, Raffaello Camoriano, Robin Hesse, Roshni Ramanna Kamath, Rupert Mitchell, Samin Hamidi, Simone Schaub-Meyer, Subarnaduti Paul, Tinne Tuytelaars, Vincenzo Lomonaco","submitted_at":"2025-02-17T15:40:13Z","abstract_excerpt":"Continual learning (CL) is the sub-field of machine learning concerned with accumulating knowledge in dynamic environments. So far, CL research has mainly focused on incremental classification tasks, where models learn to classify new categories while retaining knowledge of previously learned ones. Here, we argue that maintaining such a focus limits both theoretical development and practical applicability of CL methods. Through a detailed analysis of concrete examples - including multi-target classification, robotics with constrained output spaces, learning in continuous task domains, and high"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.11927","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/2502.11927/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":"2502.11927","created_at":"2026-07-05T10:15:36.299675+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.11927v1","created_at":"2026-07-05T10:15:36.299675+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.11927","created_at":"2026-07-05T10:15:36.299675+00:00"},{"alias_kind":"pith_short_12","alias_value":"YKRUGGKT2JX6","created_at":"2026-07-05T10:15:36.299675+00:00"},{"alias_kind":"pith_short_16","alias_value":"YKRUGGKT2JX63MZN","created_at":"2026-07-05T10:15:36.299675+00:00"},{"alias_kind":"pith_short_8","alias_value":"YKRUGGKT","created_at":"2026-07-05T10:15:36.299675+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.26097","citing_title":"Forgetting in Language Models: Capacity, Optimization, and Self-Generated Replay","ref_index":47,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB","json":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB.json","graph_json":"https://pith.science/api/pith-number/YKRUGGKT2JX63MZNVZH7ODLEKB/graph.json","events_json":"https://pith.science/api/pith-number/YKRUGGKT2JX63MZNVZH7ODLEKB/events.json","paper":"https://pith.science/paper/YKRUGGKT"},"agent_actions":{"view_html":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB","download_json":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB.json","view_paper":"https://pith.science/paper/YKRUGGKT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.11927&json=true","fetch_graph":"https://pith.science/api/pith-number/YKRUGGKT2JX63MZNVZH7ODLEKB/graph.json","fetch_events":"https://pith.science/api/pith-number/YKRUGGKT2JX63MZNVZH7ODLEKB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB/action/storage_attestation","attest_author":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB/action/author_attestation","sign_citation":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB/action/citation_signature","submit_replication":"https://pith.science/pith/YKRUGGKT2JX63MZNVZH7ODLEKB/action/replication_record"}},"created_at":"2026-07-05T10:15:36.299675+00:00","updated_at":"2026-07-05T10:15:36.299675+00:00"}