{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:LCYKJKL3XL72ROOXKI2IZD37UD","short_pith_number":"pith:LCYKJKL3","schema_version":"1.0","canonical_sha256":"58b0a4a97bbaffa8b9d752348c8f7fa0e24ee36581c5d8caad4ff162665626be","source":{"kind":"arxiv","id":"2010.07733","version":3},"attestation_state":"computed","paper":{"title":"R-GAP: Recursive Gradient Attack on Privacy","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Junyi Zhu, Matthew Blaschko","submitted_at":"2020-10-15T13:22:40Z","abstract_excerpt":"Federated learning frameworks have been regarded as a promising approach to break the dilemma between demands on privacy and the promise of learning from large collections of distributed data. Many such frameworks only ask collaborators to share their local update of a common model, i.e. gradients with respect to locally stored data, instead of exposing their raw data to other collaborators. However, recent optimization-based gradient attacks show that raw data can often be accurately recovered from gradients. It has been shown that minimizing the Euclidean distance between true gradients and "},"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":"2010.07733","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-10-15T13:22:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"671a1f659d8531d3e54d1308a2f6192ed687082c4220cb0b856b33903021d290","abstract_canon_sha256":"80e0cf85d115d19dd9a6216e0d59b76af31bf7311186d281921e1a31befe783c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:23:15.992489Z","signature_b64":"kmDbjryqbaKh5WgmkMkQkZPP4kafaw0TwoyF0JkjUtzpgIzvjBqaDGBr9u2MkUaRlVzVEBZuuLsqnUaxEsv5Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"58b0a4a97bbaffa8b9d752348c8f7fa0e24ee36581c5d8caad4ff162665626be","last_reissued_at":"2026-07-05T02:23:15.992067Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:23:15.992067Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"R-GAP: Recursive Gradient Attack on Privacy","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Junyi Zhu, Matthew Blaschko","submitted_at":"2020-10-15T13:22:40Z","abstract_excerpt":"Federated learning frameworks have been regarded as a promising approach to break the dilemma between demands on privacy and the promise of learning from large collections of distributed data. Many such frameworks only ask collaborators to share their local update of a common model, i.e. gradients with respect to locally stored data, instead of exposing their raw data to other collaborators. However, recent optimization-based gradient attacks show that raw data can often be accurately recovered from gradients. It has been shown that minimizing the Euclidean distance between true gradients and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.07733","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/2010.07733/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":"2010.07733","created_at":"2026-07-05T02:23:15.992125+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.07733v3","created_at":"2026-07-05T02:23:15.992125+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.07733","created_at":"2026-07-05T02:23:15.992125+00:00"},{"alias_kind":"pith_short_12","alias_value":"LCYKJKL3XL72","created_at":"2026-07-05T02:23:15.992125+00:00"},{"alias_kind":"pith_short_16","alias_value":"LCYKJKL3XL72ROOX","created_at":"2026-07-05T02:23:15.992125+00:00"},{"alias_kind":"pith_short_8","alias_value":"LCYKJKL3","created_at":"2026-07-05T02:23:15.992125+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20553","citing_title":"From Efficiency to Leakage -- Privacy Backdoor in Federated Language Model Fine-Tuning","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20546","citing_title":"Predictability as a Fine-Grained Measure for Privacy","ref_index":59,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06297","citing_title":"FedSpy-LLM: Towards Scalable and Generalizable Data Reconstruction Attacks from Gradients on LLMs","ref_index":44,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD","json":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD.json","graph_json":"https://pith.science/api/pith-number/LCYKJKL3XL72ROOXKI2IZD37UD/graph.json","events_json":"https://pith.science/api/pith-number/LCYKJKL3XL72ROOXKI2IZD37UD/events.json","paper":"https://pith.science/paper/LCYKJKL3"},"agent_actions":{"view_html":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD","download_json":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD.json","view_paper":"https://pith.science/paper/LCYKJKL3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.07733&json=true","fetch_graph":"https://pith.science/api/pith-number/LCYKJKL3XL72ROOXKI2IZD37UD/graph.json","fetch_events":"https://pith.science/api/pith-number/LCYKJKL3XL72ROOXKI2IZD37UD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD/action/storage_attestation","attest_author":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD/action/author_attestation","sign_citation":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD/action/citation_signature","submit_replication":"https://pith.science/pith/LCYKJKL3XL72ROOXKI2IZD37UD/action/replication_record"}},"created_at":"2026-07-05T02:23:15.992125+00:00","updated_at":"2026-07-05T02:23:15.992125+00:00"}