{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:SAVTZMBOFHLMUYPTXWZ5YQRYNB","short_pith_number":"pith:SAVTZMBO","schema_version":"1.0","canonical_sha256":"902b3cb02e29d6ca61f3bdb3dc42386873f8608152a0ec753768f44895d2f6ab","source":{"kind":"arxiv","id":"2102.13179","version":1},"attestation_state":"computed","paper":{"title":"Machine Unlearning via Algorithmic Stability","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Anup Rao, Enayat Ullah, Raman Arora, Ryan Rossi, Tung Mai","submitted_at":"2021-02-25T21:16:56Z","abstract_excerpt":"We study the problem of machine unlearning and identify a notion of algorithmic stability, Total Variation (TV) stability, which we argue, is suitable for the goal of exact unlearning. For convex risk minimization problems, we design TV-stable algorithms based on noisy Stochastic Gradient Descent (SGD). Our key contribution is the design of corresponding efficient unlearning algorithms, which are based on constructing a (maximal) coupling of Markov chains for the noisy SGD procedure. To understand the trade-offs between accuracy and unlearning efficiency, we give upper and lower bounds on exce"},"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":"2102.13179","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-02-25T21:16:56Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"45a324203bfb3fa3de83b76dc9c7d81b8d72f66bc9c9c95990dad03d6e7ff67e","abstract_canon_sha256":"d10a696dae36edf5462fc9cc35f963ce6f80868f161a9a1ad9bb727725f7096a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:18:34.787479Z","signature_b64":"onWWOFtlvyDDQPsaFds3etSs70Sb64QWiGoPuby4n9RJFgCxyYy3yYg4EQI/xW9QnKMDql1bgI5PfxqAlj+3AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"902b3cb02e29d6ca61f3bdb3dc42386873f8608152a0ec753768f44895d2f6ab","last_reissued_at":"2026-07-05T02:18:34.786959Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:18:34.786959Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Machine Unlearning via Algorithmic Stability","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Anup Rao, Enayat Ullah, Raman Arora, Ryan Rossi, Tung Mai","submitted_at":"2021-02-25T21:16:56Z","abstract_excerpt":"We study the problem of machine unlearning and identify a notion of algorithmic stability, Total Variation (TV) stability, which we argue, is suitable for the goal of exact unlearning. For convex risk minimization problems, we design TV-stable algorithms based on noisy Stochastic Gradient Descent (SGD). Our key contribution is the design of corresponding efficient unlearning algorithms, which are based on constructing a (maximal) coupling of Markov chains for the noisy SGD procedure. To understand the trade-offs between accuracy and unlearning efficiency, we give upper and lower bounds on exce"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.13179","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/2102.13179/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":"2102.13179","created_at":"2026-07-05T02:18:34.787029+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.13179v1","created_at":"2026-07-05T02:18:34.787029+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.13179","created_at":"2026-07-05T02:18:34.787029+00:00"},{"alias_kind":"pith_short_12","alias_value":"SAVTZMBOFHLM","created_at":"2026-07-05T02:18:34.787029+00:00"},{"alias_kind":"pith_short_16","alias_value":"SAVTZMBOFHLMUYPT","created_at":"2026-07-05T02:18:34.787029+00:00"},{"alias_kind":"pith_short_8","alias_value":"SAVTZMBO","created_at":"2026-07-05T02:18:34.787029+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.23046","citing_title":"Shape of Memory: a Geometric Analysis of Machine Unlearning in Second-Order Optimizers","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB","json":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB.json","graph_json":"https://pith.science/api/pith-number/SAVTZMBOFHLMUYPTXWZ5YQRYNB/graph.json","events_json":"https://pith.science/api/pith-number/SAVTZMBOFHLMUYPTXWZ5YQRYNB/events.json","paper":"https://pith.science/paper/SAVTZMBO"},"agent_actions":{"view_html":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB","download_json":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB.json","view_paper":"https://pith.science/paper/SAVTZMBO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.13179&json=true","fetch_graph":"https://pith.science/api/pith-number/SAVTZMBOFHLMUYPTXWZ5YQRYNB/graph.json","fetch_events":"https://pith.science/api/pith-number/SAVTZMBOFHLMUYPTXWZ5YQRYNB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB/action/storage_attestation","attest_author":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB/action/author_attestation","sign_citation":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB/action/citation_signature","submit_replication":"https://pith.science/pith/SAVTZMBOFHLMUYPTXWZ5YQRYNB/action/replication_record"}},"created_at":"2026-07-05T02:18:34.787029+00:00","updated_at":"2026-07-05T02:18:34.787029+00:00"}