{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:M4H3OYQIIGAADIBECA3NGFYWLB","short_pith_number":"pith:M4H3OYQI","schema_version":"1.0","canonical_sha256":"670fb76208418001a0241036d31716585037830d37f5761a61dd3eabec6df7be","source":{"kind":"arxiv","id":"2503.18985","version":2},"attestation_state":"computed","paper":{"title":"LoRA Subtraction for Drift-Resistant Space in Exemplar-Free Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Xiaobin Chang, Xuan Liu","submitted_at":"2025-03-23T07:38:53Z","abstract_excerpt":"In continual learning (CL), catastrophic forgetting often arises due to feature drift. This challenge is particularly prominent in the exemplar-free continual learning (EFCL) setting, where samples from previous tasks cannot be retained, making it difficult to preserve prior knowledge. To address this issue, some EFCL methods aim to identify feature spaces that minimize the impact on previous tasks while accommodating new ones. However, they rely on static features or outdated statistics stored from old tasks, which prevents them from capturing the dynamic evolution of the feature space in CL,"},"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":"2503.18985","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-03-23T07:38:53Z","cross_cats_sorted":["cs.AI","cs.CV","stat.ML"],"title_canon_sha256":"a5f55c85e6831204519f05fe766440054bcd8b6f6574a538c31ed8d78304b3f6","abstract_canon_sha256":"24f8525ade9af8e8e6d14ed72c4fdf5fed5bf045f14c29294bc623064827d134"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:41:53.994168Z","signature_b64":"hoCU4ID28SsPofmut4YyVd4K3rHIGfSNC0J4qP4bI482hsiN7pg6I/gV2VoYNuHpZ7YqUmWvzCoYjj5xj/H7Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"670fb76208418001a0241036d31716585037830d37f5761a61dd3eabec6df7be","last_reissued_at":"2026-07-05T10:41:53.993698Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:41:53.993698Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LoRA Subtraction for Drift-Resistant Space in Exemplar-Free Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Xiaobin Chang, Xuan Liu","submitted_at":"2025-03-23T07:38:53Z","abstract_excerpt":"In continual learning (CL), catastrophic forgetting often arises due to feature drift. This challenge is particularly prominent in the exemplar-free continual learning (EFCL) setting, where samples from previous tasks cannot be retained, making it difficult to preserve prior knowledge. To address this issue, some EFCL methods aim to identify feature spaces that minimize the impact on previous tasks while accommodating new ones. However, they rely on static features or outdated statistics stored from old tasks, which prevents them from capturing the dynamic evolution of the feature space in CL,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.18985","kind":"arxiv","version":2},"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/2503.18985/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":"2503.18985","created_at":"2026-07-05T10:41:53.993757+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.18985v2","created_at":"2026-07-05T10:41:53.993757+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.18985","created_at":"2026-07-05T10:41:53.993757+00:00"},{"alias_kind":"pith_short_12","alias_value":"M4H3OYQIIGAA","created_at":"2026-07-05T10:41:53.993757+00:00"},{"alias_kind":"pith_short_16","alias_value":"M4H3OYQIIGAADIBE","created_at":"2026-07-05T10:41:53.993757+00:00"},{"alias_kind":"pith_short_8","alias_value":"M4H3OYQI","created_at":"2026-07-05T10:41:53.993757+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.30190","citing_title":"Few-Shot Domain Incremental Learning via Continual Vision-Language Consolidation","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2508.18187","citing_title":"BRAIN: Bias-Mitigation Continual Learning Approach to Vision-Brain Understanding","ref_index":87,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB","json":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB.json","graph_json":"https://pith.science/api/pith-number/M4H3OYQIIGAADIBECA3NGFYWLB/graph.json","events_json":"https://pith.science/api/pith-number/M4H3OYQIIGAADIBECA3NGFYWLB/events.json","paper":"https://pith.science/paper/M4H3OYQI"},"agent_actions":{"view_html":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB","download_json":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB.json","view_paper":"https://pith.science/paper/M4H3OYQI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.18985&json=true","fetch_graph":"https://pith.science/api/pith-number/M4H3OYQIIGAADIBECA3NGFYWLB/graph.json","fetch_events":"https://pith.science/api/pith-number/M4H3OYQIIGAADIBECA3NGFYWLB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB/action/storage_attestation","attest_author":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB/action/author_attestation","sign_citation":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB/action/citation_signature","submit_replication":"https://pith.science/pith/M4H3OYQIIGAADIBECA3NGFYWLB/action/replication_record"}},"created_at":"2026-07-05T10:41:53.993757+00:00","updated_at":"2026-07-05T10:41:53.993757+00:00"}