{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:OI6JVRSN6AZVWFYIIRKJO55Y62","short_pith_number":"pith:OI6JVRSN","schema_version":"1.0","canonical_sha256":"723c9ac64df0335b170844549777b8f6854a89bbb2e50fee4cff404e395774cf","source":{"kind":"arxiv","id":"2310.05331","version":1},"attestation_state":"computed","paper":{"title":"Unlearning with Fisher Masking","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aimin Zhou, Changzhi Sun, Yuanbin Wu, Yufang Liu","submitted_at":"2023-10-09T01:24:06Z","abstract_excerpt":"Machine unlearning aims to revoke some training data after learning in response to requests from users, model developers, and administrators. Most previous methods are based on direct fine-tuning, which may neither remove data completely nor retain full performances on the remain data. In this work, we find that, by first masking some important parameters before fine-tuning, the performances of unlearning could be significantly improved. We propose a new masking strategy tailored to unlearning based on Fisher information. Experiments on various datasets and network structures show the effectiv"},"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":"2310.05331","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-10-09T01:24:06Z","cross_cats_sorted":[],"title_canon_sha256":"f35d66e4f36d7746f5be4a2fb7a525e08cae87ca5cf8c2168eb47b3574b76d38","abstract_canon_sha256":"d2c8452cc20a9b2c10646be26a7d4bbd8371161e7fd38b9a46e43b6c9a1e8756"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:58:43.832164Z","signature_b64":"6mxxjK+xdhhDdbIdf7qYb+UL8bvfhZ4N24KgmdfGKU/NLYyoLj8B0JHYC18OyDW43LijsXiiA/qOrtZUeUJ1CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"723c9ac64df0335b170844549777b8f6854a89bbb2e50fee4cff404e395774cf","last_reissued_at":"2026-07-05T06:58:43.831657Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:58:43.831657Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Unlearning with Fisher Masking","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aimin Zhou, Changzhi Sun, Yuanbin Wu, Yufang Liu","submitted_at":"2023-10-09T01:24:06Z","abstract_excerpt":"Machine unlearning aims to revoke some training data after learning in response to requests from users, model developers, and administrators. Most previous methods are based on direct fine-tuning, which may neither remove data completely nor retain full performances on the remain data. In this work, we find that, by first masking some important parameters before fine-tuning, the performances of unlearning could be significantly improved. We propose a new masking strategy tailored to unlearning based on Fisher information. Experiments on various datasets and network structures show the effectiv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.05331","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/2310.05331/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":"2310.05331","created_at":"2026-07-05T06:58:43.831722+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.05331v1","created_at":"2026-07-05T06:58:43.831722+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.05331","created_at":"2026-07-05T06:58:43.831722+00:00"},{"alias_kind":"pith_short_12","alias_value":"OI6JVRSN6AZV","created_at":"2026-07-05T06:58:43.831722+00:00"},{"alias_kind":"pith_short_16","alias_value":"OI6JVRSN6AZVWFYI","created_at":"2026-07-05T06:58:43.831722+00:00"},{"alias_kind":"pith_short_8","alias_value":"OI6JVRSN","created_at":"2026-07-05T06:58:43.831722+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2505.10859","citing_title":"Exploring Nonlinear Pathway in Parameter Space for Machine Unlearning","ref_index":20,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62","json":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62.json","graph_json":"https://pith.science/api/pith-number/OI6JVRSN6AZVWFYIIRKJO55Y62/graph.json","events_json":"https://pith.science/api/pith-number/OI6JVRSN6AZVWFYIIRKJO55Y62/events.json","paper":"https://pith.science/paper/OI6JVRSN"},"agent_actions":{"view_html":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62","download_json":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62.json","view_paper":"https://pith.science/paper/OI6JVRSN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.05331&json=true","fetch_graph":"https://pith.science/api/pith-number/OI6JVRSN6AZVWFYIIRKJO55Y62/graph.json","fetch_events":"https://pith.science/api/pith-number/OI6JVRSN6AZVWFYIIRKJO55Y62/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62/action/storage_attestation","attest_author":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62/action/author_attestation","sign_citation":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62/action/citation_signature","submit_replication":"https://pith.science/pith/OI6JVRSN6AZVWFYIIRKJO55Y62/action/replication_record"}},"created_at":"2026-07-05T06:58:43.831722+00:00","updated_at":"2026-07-05T06:58:43.831722+00:00"}