{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:L74ZRJLMZ7ETURA3XFAMIP3M6T","short_pith_number":"pith:L74ZRJLM","schema_version":"1.0","canonical_sha256":"5ff998a56ccfc93a441bb940c43f6cf4c86d8857451539a2a44423b37f67d506","source":{"kind":"arxiv","id":"2007.11524","version":1},"attestation_state":"computed","paper":{"title":"Improving Deep Learning with Differential Privacy using Gradient Encoding and Denoising","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR","stat.ML"],"primary_cat":"cs.LG","authors_text":"Amir Houmansadr, Milad Nasr, Reza Shokri","submitted_at":"2020-07-22T16:33:14Z","abstract_excerpt":"Deep learning models leak significant amounts of information about their training datasets. Previous work has investigated training models with differential privacy (DP) guarantees through adding DP noise to the gradients. However, such solutions (specifically, DPSGD), result in large degradations in the accuracy of the trained models. In this paper, we aim at training deep learning models with DP guarantees while preserving model accuracy much better than previous works. Our key technique is to encode gradients to map them to a smaller vector space, therefore enabling us to obtain DP guarante"},"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":"2007.11524","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-07-22T16:33:14Z","cross_cats_sorted":["cs.CR","stat.ML"],"title_canon_sha256":"7d6244e29b81f2be8d8c49b839e07b823257b671b6067475a89b789d01cd6dfb","abstract_canon_sha256":"30beb989ba8267cb3263029b5df39a7323322bcc54dbeede4d27dabc7f16fbef"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:21:26.960648Z","signature_b64":"vJuwlX0o9gPsFV1D5m3etxuhSb0cgGfbs1FkcJPA4/WDmXR6KFIne6J64LVxMdOiPNN2lk5jlorLhHIK4ajxDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5ff998a56ccfc93a441bb940c43f6cf4c86d8857451539a2a44423b37f67d506","last_reissued_at":"2026-07-05T01:21:26.960262Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:21:26.960262Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Deep Learning with Differential Privacy using Gradient Encoding and Denoising","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR","stat.ML"],"primary_cat":"cs.LG","authors_text":"Amir Houmansadr, Milad Nasr, Reza Shokri","submitted_at":"2020-07-22T16:33:14Z","abstract_excerpt":"Deep learning models leak significant amounts of information about their training datasets. Previous work has investigated training models with differential privacy (DP) guarantees through adding DP noise to the gradients. However, such solutions (specifically, DPSGD), result in large degradations in the accuracy of the trained models. In this paper, we aim at training deep learning models with DP guarantees while preserving model accuracy much better than previous works. Our key technique is to encode gradients to map them to a smaller vector space, therefore enabling us to obtain DP guarante"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2007.11524","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/2007.11524/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":"2007.11524","created_at":"2026-07-05T01:21:26.960323+00:00"},{"alias_kind":"arxiv_version","alias_value":"2007.11524v1","created_at":"2026-07-05T01:21:26.960323+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2007.11524","created_at":"2026-07-05T01:21:26.960323+00:00"},{"alias_kind":"pith_short_12","alias_value":"L74ZRJLMZ7ET","created_at":"2026-07-05T01:21:26.960323+00:00"},{"alias_kind":"pith_short_16","alias_value":"L74ZRJLMZ7ETURA3","created_at":"2026-07-05T01:21:26.960323+00:00"},{"alias_kind":"pith_short_8","alias_value":"L74ZRJLM","created_at":"2026-07-05T01:21:26.960323+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21107","citing_title":"Enhancing Differentially Private Mechanisms via Empirical Bayes","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T","json":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T.json","graph_json":"https://pith.science/api/pith-number/L74ZRJLMZ7ETURA3XFAMIP3M6T/graph.json","events_json":"https://pith.science/api/pith-number/L74ZRJLMZ7ETURA3XFAMIP3M6T/events.json","paper":"https://pith.science/paper/L74ZRJLM"},"agent_actions":{"view_html":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T","download_json":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T.json","view_paper":"https://pith.science/paper/L74ZRJLM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2007.11524&json=true","fetch_graph":"https://pith.science/api/pith-number/L74ZRJLMZ7ETURA3XFAMIP3M6T/graph.json","fetch_events":"https://pith.science/api/pith-number/L74ZRJLMZ7ETURA3XFAMIP3M6T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T/action/storage_attestation","attest_author":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T/action/author_attestation","sign_citation":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T/action/citation_signature","submit_replication":"https://pith.science/pith/L74ZRJLMZ7ETURA3XFAMIP3M6T/action/replication_record"}},"created_at":"2026-07-05T01:21:26.960323+00:00","updated_at":"2026-07-05T01:21:26.960323+00:00"}