{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:U3BLCKNZ5KROXNYIBY32ZAUUMS","short_pith_number":"pith:U3BLCKNZ","schema_version":"1.0","canonical_sha256":"a6c2b129b9eaa2ebb7080e37ac829464bbb63b84d33a374af5ae8ad10a0e6494","source":{"kind":"arxiv","id":"2311.14632","version":2},"attestation_state":"computed","paper":{"title":"Differentially Private SGD Without Clipping Bias: An Error-Feedback Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Mingyi Hong, Xinwei Zhang, Zhiqi Bu, Zhiwei Steven Wu","submitted_at":"2023-11-24T17:56:44Z","abstract_excerpt":"Differentially Private Stochastic Gradient Descent with Gradient Clipping (DPSGD-GC) is a powerful tool for training deep learning models using sensitive data, providing both a solid theoretical privacy guarantee and high efficiency. However, using DPSGD-GC to ensure Differential Privacy (DP) comes at the cost of model performance degradation due to DP noise injection and gradient clipping. Existing research has extensively analyzed the theoretical convergence of DPSGD-GC, and has shown that it only converges when using large clipping thresholds that are dependent on problem-specific parameter"},"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":"2311.14632","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-11-24T17:56:44Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"7d3cd4db92be820692830bd14d154abfbfdb686cc5ec3732be6f5ccc30b3b447","abstract_canon_sha256":"459a36f89c8f0b90c2853b467473292b5a01d8e9b0e7be9e0d59da40cafa07aa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:08:53.035416Z","signature_b64":"R/ltVy3A8L7I4aUNQMhHMVEddnHHoGfrIIoyz+v9FFbbnoCGdpEGpj9P/egtoVMFephYHsffoy55Vdi6MKLmDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a6c2b129b9eaa2ebb7080e37ac829464bbb63b84d33a374af5ae8ad10a0e6494","last_reissued_at":"2026-07-05T08:08:53.035043Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:08:53.035043Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Differentially Private SGD Without Clipping Bias: An Error-Feedback Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Mingyi Hong, Xinwei Zhang, Zhiqi Bu, Zhiwei Steven Wu","submitted_at":"2023-11-24T17:56:44Z","abstract_excerpt":"Differentially Private Stochastic Gradient Descent with Gradient Clipping (DPSGD-GC) is a powerful tool for training deep learning models using sensitive data, providing both a solid theoretical privacy guarantee and high efficiency. However, using DPSGD-GC to ensure Differential Privacy (DP) comes at the cost of model performance degradation due to DP noise injection and gradient clipping. Existing research has extensively analyzed the theoretical convergence of DPSGD-GC, and has shown that it only converges when using large clipping thresholds that are dependent on problem-specific parameter"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.14632","kind":"arxiv","version":2},"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/2311.14632/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":"2311.14632","created_at":"2026-07-05T08:08:53.035100+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.14632v2","created_at":"2026-07-05T08:08:53.035100+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.14632","created_at":"2026-07-05T08:08:53.035100+00:00"},{"alias_kind":"pith_short_12","alias_value":"U3BLCKNZ5KRO","created_at":"2026-07-05T08:08:53.035100+00:00"},{"alias_kind":"pith_short_16","alias_value":"U3BLCKNZ5KROXNYI","created_at":"2026-07-05T08:08:53.035100+00:00"},{"alias_kind":"pith_short_8","alias_value":"U3BLCKNZ","created_at":"2026-07-05T08:08:53.035100+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/U3BLCKNZ5KROXNYIBY32ZAUUMS","json":"https://pith.science/pith/U3BLCKNZ5KROXNYIBY32ZAUUMS.json","graph_json":"https://pith.science/api/pith-number/U3BLCKNZ5KROXNYIBY32ZAUUMS/graph.json","events_json":"https://pith.science/api/pith-number/U3BLCKNZ5KROXNYIBY32ZAUUMS/events.json","paper":"https://pith.science/paper/U3BLCKNZ"},"agent_actions":{"view_html":"https://pith.science/pith/U3BLCKNZ5KROXNYIBY32ZAUUMS","download_json":"https://pith.science/pith/U3BLCKNZ5KROXNYIBY32ZAUUMS.json","view_paper":"https://pith.science/paper/U3BLCKNZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.14632&json=true","fetch_graph":"https://pith.science/api/pith-number/U3BLCKNZ5KROXNYIBY32ZAUUMS/graph.json","fetch_events":"https://pith.science/api/pith-number/U3BLCKNZ5KROXNYIBY32ZAUUMS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/U3BLCKNZ5KROXNYIBY32ZAUUMS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/U3BLCKNZ5KROXNYIBY32ZAUUMS/action/storage_attestation","attest_author":"https://pith.science/pith/U3BLCKNZ5KROXNYIBY32ZAUUMS/action/author_attestation","sign_citation":"https://pith.science/pith/U3BLCKNZ5KROXNYIBY32ZAUUMS/action/citation_signature","submit_replication":"https://pith.science/pith/U3BLCKNZ5KROXNYIBY32ZAUUMS/action/replication_record"}},"created_at":"2026-07-05T08:08:53.035100+00:00","updated_at":"2026-07-05T08:08:53.035100+00:00"}