{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:5QY3FD737GWLWFCNDB7PYJVBPB","short_pith_number":"pith:5QY3FD73","schema_version":"1.0","canonical_sha256":"ec31b28ffbf9acbb144d187efc26a178466f2ad71e5716954f7e66f962d6e7cf","source":{"kind":"arxiv","id":"2305.08846","version":1},"attestation_state":"computed","paper":{"title":"Privacy Auditing with One (1) Training Run","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR","cs.DS"],"primary_cat":"cs.LG","authors_text":"Matthew Jagielski, Milad Nasr, Thomas Steinke","submitted_at":"2023-05-15T17:57:56Z","abstract_excerpt":"We propose a scheme for auditing differentially private machine learning systems with a single training run. This exploits the parallelism of being able to add or remove multiple training examples independently. We analyze this using the connection between differential privacy and statistical generalization, which avoids the cost of group privacy. Our auditing scheme requires minimal assumptions about the algorithm and can be applied in the black-box or white-box setting."},"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":"2305.08846","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-05-15T17:57:56Z","cross_cats_sorted":["cs.CR","cs.DS"],"title_canon_sha256":"f25f81b537cd84761ca2dfa8bed5e1e705118b1873a2b9b4131d3dd7f8742472","abstract_canon_sha256":"ec14592b2bff5b3a2efe22569dceb8ccf50235b777ae5dc616726fbe107ba43b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:10:12.229170Z","signature_b64":"uaclXI6gxek19992q4cn+H3LBfq0sHxiKs6PLg7absWXCwm68e3qHwOKBSreUafO5JrAvqYzD2BSA1Cd3fI4CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ec31b28ffbf9acbb144d187efc26a178466f2ad71e5716954f7e66f962d6e7cf","last_reissued_at":"2026-07-05T06:10:12.228744Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:10:12.228744Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Privacy Auditing with One (1) Training Run","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR","cs.DS"],"primary_cat":"cs.LG","authors_text":"Matthew Jagielski, Milad Nasr, Thomas Steinke","submitted_at":"2023-05-15T17:57:56Z","abstract_excerpt":"We propose a scheme for auditing differentially private machine learning systems with a single training run. This exploits the parallelism of being able to add or remove multiple training examples independently. We analyze this using the connection between differential privacy and statistical generalization, which avoids the cost of group privacy. Our auditing scheme requires minimal assumptions about the algorithm and can be applied in the black-box or white-box setting."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.08846","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/2305.08846/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":"2305.08846","created_at":"2026-07-05T06:10:12.228795+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.08846v1","created_at":"2026-07-05T06:10:12.228795+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.08846","created_at":"2026-07-05T06:10:12.228795+00:00"},{"alias_kind":"pith_short_12","alias_value":"5QY3FD737GWL","created_at":"2026-07-05T06:10:12.228795+00:00"},{"alias_kind":"pith_short_16","alias_value":"5QY3FD737GWLWFCN","created_at":"2026-07-05T06:10:12.228795+00:00"},{"alias_kind":"pith_short_8","alias_value":"5QY3FD73","created_at":"2026-07-05T06:10:12.228795+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.05898","citing_title":"Auditing of Unlearning Algorithms","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2606.26627","citing_title":"Agents That Know Too Much: A Data-Centric Survey of Privacy in LLM Agents","ref_index":107,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15648","citing_title":"Rethinking the Security of DP-SGD: A Corrected Analysis of Differentially Private Machine Learning","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2310.16789","citing_title":"Detecting Pretraining Data from Large Language Models","ref_index":108,"is_internal_anchor":false},{"citing_arxiv_id":"2401.06121","citing_title":"TOFU: A Task of Fictitious Unlearning for LLMs","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11170","citing_title":"Unlearning with Asymmetric Sources: Improved Unlearning-Utility Trade-off with Public Data","ref_index":49,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB","json":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB.json","graph_json":"https://pith.science/api/pith-number/5QY3FD737GWLWFCNDB7PYJVBPB/graph.json","events_json":"https://pith.science/api/pith-number/5QY3FD737GWLWFCNDB7PYJVBPB/events.json","paper":"https://pith.science/paper/5QY3FD73"},"agent_actions":{"view_html":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB","download_json":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB.json","view_paper":"https://pith.science/paper/5QY3FD73","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.08846&json=true","fetch_graph":"https://pith.science/api/pith-number/5QY3FD737GWLWFCNDB7PYJVBPB/graph.json","fetch_events":"https://pith.science/api/pith-number/5QY3FD737GWLWFCNDB7PYJVBPB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB/action/storage_attestation","attest_author":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB/action/author_attestation","sign_citation":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB/action/citation_signature","submit_replication":"https://pith.science/pith/5QY3FD737GWLWFCNDB7PYJVBPB/action/replication_record"}},"created_at":"2026-07-05T06:10:12.228795+00:00","updated_at":"2026-07-05T06:10:12.228795+00:00"}