{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:LYQ3HQZWROJZ55ZJD7T3MEMYKT","short_pith_number":"pith:LYQ3HQZW","schema_version":"1.0","canonical_sha256":"5e21b3c3368b939ef7291fe7b6119854d1c7700cb00263d81378e9fa7d0cea69","source":{"kind":"arxiv","id":"2102.11673","version":3},"attestation_state":"computed","paper":{"title":"Measuring Data Leakage in Machine-Learning Models with Fisher Information","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Awni Hannun, Chuan Guo, Laurens van der Maaten","submitted_at":"2021-02-23T13:02:34Z","abstract_excerpt":"Machine-learning models contain information about the data they were trained on. This information leaks either through the model itself or through predictions made by the model. Consequently, when the training data contains sensitive attributes, assessing the amount of information leakage is paramount. We propose a method to quantify this leakage using the Fisher information of the model about the data. Unlike the worst-case a priori guarantees of differential privacy, Fisher information loss measures leakage with respect to specific examples, attributes, or sub-populations within the dataset."},"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":"2102.11673","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-02-23T13:02:34Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"6e923521fd72cb17d4e3ed225067b046d363fcd480af0cc1f6952ef6198e7aa8","abstract_canon_sha256":"36fc9a229b5a3ce8ca6262dad4a87c1ab7e9512ac9e6d105f3ec468ad7ccc621"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:08:25.574143Z","signature_b64":"gux2e9KdNhTHZFlOGRvyYeM7bMK27QJJ7/PVQWxnJcgRxdP3nbu1dQHfQdvnVS4Sr6nyi5rp4CbCdTaT6tBeAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5e21b3c3368b939ef7291fe7b6119854d1c7700cb00263d81378e9fa7d0cea69","last_reissued_at":"2026-07-05T03:08:25.573704Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:08:25.573704Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Measuring Data Leakage in Machine-Learning Models with Fisher Information","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Awni Hannun, Chuan Guo, Laurens van der Maaten","submitted_at":"2021-02-23T13:02:34Z","abstract_excerpt":"Machine-learning models contain information about the data they were trained on. This information leaks either through the model itself or through predictions made by the model. Consequently, when the training data contains sensitive attributes, assessing the amount of information leakage is paramount. We propose a method to quantify this leakage using the Fisher information of the model about the data. Unlike the worst-case a priori guarantees of differential privacy, Fisher information loss measures leakage with respect to specific examples, attributes, or sub-populations within the dataset."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.11673","kind":"arxiv","version":3},"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/2102.11673/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":"2102.11673","created_at":"2026-07-05T03:08:25.573765+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.11673v3","created_at":"2026-07-05T03:08:25.573765+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.11673","created_at":"2026-07-05T03:08:25.573765+00:00"},{"alias_kind":"pith_short_12","alias_value":"LYQ3HQZWROJZ","created_at":"2026-07-05T03:08:25.573765+00:00"},{"alias_kind":"pith_short_16","alias_value":"LYQ3HQZWROJZ55ZJ","created_at":"2026-07-05T03:08:25.573765+00:00"},{"alias_kind":"pith_short_8","alias_value":"LYQ3HQZW","created_at":"2026-07-05T03:08:25.573765+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT","json":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT.json","graph_json":"https://pith.science/api/pith-number/LYQ3HQZWROJZ55ZJD7T3MEMYKT/graph.json","events_json":"https://pith.science/api/pith-number/LYQ3HQZWROJZ55ZJD7T3MEMYKT/events.json","paper":"https://pith.science/paper/LYQ3HQZW"},"agent_actions":{"view_html":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT","download_json":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT.json","view_paper":"https://pith.science/paper/LYQ3HQZW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.11673&json=true","fetch_graph":"https://pith.science/api/pith-number/LYQ3HQZWROJZ55ZJD7T3MEMYKT/graph.json","fetch_events":"https://pith.science/api/pith-number/LYQ3HQZWROJZ55ZJD7T3MEMYKT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT/action/storage_attestation","attest_author":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT/action/author_attestation","sign_citation":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT/action/citation_signature","submit_replication":"https://pith.science/pith/LYQ3HQZWROJZ55ZJD7T3MEMYKT/action/replication_record"}},"created_at":"2026-07-05T03:08:25.573765+00:00","updated_at":"2026-07-05T03:08:25.573765+00:00"}