{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:FST2K7CTN74JBMFWUM4FS5UPJV","short_pith_number":"pith:FST2K7CT","schema_version":"1.0","canonical_sha256":"2ca7a57c536ff890b0b6a33859768f4d74bd6e5b9073dd0937ef0b06ee78be4d","source":{"kind":"arxiv","id":"2506.08201","version":1},"attestation_state":"computed","paper":{"title":"Correlated Noise Mechanisms for Differentially Private Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Abhradeep Thakurta, Arun Ganesh, Christopher A. Choquette-Choo, H. Brendan McMahan, Jalaj Upadhyay, Jonathan Katz, Keith Rush, Krishnamurthy Dvijotham, Krishna Pillutla, Monika Henzinger, Ryan McKenna, Thomas Steinke","submitted_at":"2025-06-09T20:21:50Z","abstract_excerpt":"This monograph explores the design and analysis of correlated noise mechanisms for differential privacy (DP), focusing on their application to private training of AI and machine learning models via the core primitive of estimation of weighted prefix sums. While typical DP mechanisms inject independent noise into each step of a stochastic gradient (SGD) learning algorithm in order to protect the privacy of the training data, a growing body of recent research demonstrates that introducing (anti-)correlations in the noise can significantly improve privacy-utility trade-offs by carefully canceling"},"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":"2506.08201","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-06-09T20:21:50Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"b9631bd16212577aae103210851a8a8ebd75ecd5fe41b08cd445a531f8965d6d","abstract_canon_sha256":"0b9815e75ced69097ffdba7ec8fc248b4880b4616538e152a421829724bba4d5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:19:09.018673Z","signature_b64":"L3wflu3NIjCUf42w+ZG/vBBJPdCXlitAj5g3QH55wj2eUHQrgtyHCfgIL8QSGENEIb7iDYl7h+fmeOiwwBfiBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2ca7a57c536ff890b0b6a33859768f4d74bd6e5b9073dd0937ef0b06ee78be4d","last_reissued_at":"2026-07-05T11:19:09.018227Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:19:09.018227Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Correlated Noise Mechanisms for Differentially Private Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Abhradeep Thakurta, Arun Ganesh, Christopher A. Choquette-Choo, H. Brendan McMahan, Jalaj Upadhyay, Jonathan Katz, Keith Rush, Krishnamurthy Dvijotham, Krishna Pillutla, Monika Henzinger, Ryan McKenna, Thomas Steinke","submitted_at":"2025-06-09T20:21:50Z","abstract_excerpt":"This monograph explores the design and analysis of correlated noise mechanisms for differential privacy (DP), focusing on their application to private training of AI and machine learning models via the core primitive of estimation of weighted prefix sums. While typical DP mechanisms inject independent noise into each step of a stochastic gradient (SGD) learning algorithm in order to protect the privacy of the training data, a growing body of recent research demonstrates that introducing (anti-)correlations in the noise can significantly improve privacy-utility trade-offs by carefully canceling"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.08201","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/2506.08201/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":"2506.08201","created_at":"2026-07-05T11:19:09.018278+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.08201v1","created_at":"2026-07-05T11:19:09.018278+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.08201","created_at":"2026-07-05T11:19:09.018278+00:00"},{"alias_kind":"pith_short_12","alias_value":"FST2K7CTN74J","created_at":"2026-07-05T11:19:09.018278+00:00"},{"alias_kind":"pith_short_16","alias_value":"FST2K7CTN74JBMFW","created_at":"2026-07-05T11:19:09.018278+00:00"},{"alias_kind":"pith_short_8","alias_value":"FST2K7CT","created_at":"2026-07-05T11:19:09.018278+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.26222","citing_title":"From Privacy to Generalization: Linear Max-Information Bounds for DP-SGD","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30476","citing_title":"Local Differential Privacy with Correlated Noise Achieves Central-DP Optimal Cost","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12648","citing_title":"Population Risk Bounds for Kolmogorov-Arnold Networks Trained by DP-SGD with Correlated Noise","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13503","citing_title":"Limits of Personalizing Differential Privacy Budgets","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15596","citing_title":"Privacy, Prediction, and Allocation","ref_index":58,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02886","citing_title":"CityOS: Privacy Architecture for Urban Sensing","ref_index":67,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV","json":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV.json","graph_json":"https://pith.science/api/pith-number/FST2K7CTN74JBMFWUM4FS5UPJV/graph.json","events_json":"https://pith.science/api/pith-number/FST2K7CTN74JBMFWUM4FS5UPJV/events.json","paper":"https://pith.science/paper/FST2K7CT"},"agent_actions":{"view_html":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV","download_json":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV.json","view_paper":"https://pith.science/paper/FST2K7CT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.08201&json=true","fetch_graph":"https://pith.science/api/pith-number/FST2K7CTN74JBMFWUM4FS5UPJV/graph.json","fetch_events":"https://pith.science/api/pith-number/FST2K7CTN74JBMFWUM4FS5UPJV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV/action/storage_attestation","attest_author":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV/action/author_attestation","sign_citation":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV/action/citation_signature","submit_replication":"https://pith.science/pith/FST2K7CTN74JBMFWUM4FS5UPJV/action/replication_record"}},"created_at":"2026-07-05T11:19:09.018278+00:00","updated_at":"2026-07-05T11:19:09.018278+00:00"}