{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:PFFUIBJ7IMFUDHMUXRDKNEUT2M","short_pith_number":"pith:PFFUIBJ7","schema_version":"1.0","canonical_sha256":"794b44053f430b419d94bc46a69293d30903e8dd9937c6011b7ddb5e9f4c7cfc","source":{"kind":"arxiv","id":"2407.05285","version":4},"attestation_state":"computed","paper":{"title":"Mjolnir: Breaking the Shield of Perturbation-Protected Gradients via Adaptive Diffusion","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Kaiwei Lin, Qihua Zhou, Ruibin Li, Siqi Cai, Song Guo, Xuan Liu","submitted_at":"2024-07-07T07:06:49Z","abstract_excerpt":"Perturbation-based mechanisms, such as differential privacy, mitigate gradient leakage attacks by introducing noise into the gradients, thereby preventing attackers from reconstructing clients' private data from the leaked gradients. However, can gradient perturbation protection mechanisms truly defend against all gradient leakage attacks? In this paper, we present the first attempt to break the shield of gradient perturbation protection in Federated Learning for the extraction of private information. We focus on common noise distributions, specifically Gaussian and Laplace, and apply our appr"},"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":"2407.05285","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-07-07T07:06:49Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"850131af7fd743a27c15f975b1325ee80f35da2e4bec7aca3586a2b49377338a","abstract_canon_sha256":"ffe576f49c257d77dd2a7b7b265fa99d32f6e2e28f852db2eabf950418b09e68"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:57:20.849247Z","signature_b64":"hZxadfKnP/OJMy0CnWILsYb+npPT9/ME59D1kHBSRt5Rqpvc41D7ILhPT7aWEX6vFICEnegmT/HZyvwCO6q5CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"794b44053f430b419d94bc46a69293d30903e8dd9937c6011b7ddb5e9f4c7cfc","last_reissued_at":"2026-07-05T09:57:20.848814Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:57:20.848814Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mjolnir: Breaking the Shield of Perturbation-Protected Gradients via Adaptive Diffusion","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Kaiwei Lin, Qihua Zhou, Ruibin Li, Siqi Cai, Song Guo, Xuan Liu","submitted_at":"2024-07-07T07:06:49Z","abstract_excerpt":"Perturbation-based mechanisms, such as differential privacy, mitigate gradient leakage attacks by introducing noise into the gradients, thereby preventing attackers from reconstructing clients' private data from the leaked gradients. However, can gradient perturbation protection mechanisms truly defend against all gradient leakage attacks? In this paper, we present the first attempt to break the shield of gradient perturbation protection in Federated Learning for the extraction of private information. We focus on common noise distributions, specifically Gaussian and Laplace, and apply our appr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.05285","kind":"arxiv","version":4},"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/2407.05285/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":"2407.05285","created_at":"2026-07-05T09:57:20.848872+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.05285v4","created_at":"2026-07-05T09:57:20.848872+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.05285","created_at":"2026-07-05T09:57:20.848872+00:00"},{"alias_kind":"pith_short_12","alias_value":"PFFUIBJ7IMFU","created_at":"2026-07-05T09:57:20.848872+00:00"},{"alias_kind":"pith_short_16","alias_value":"PFFUIBJ7IMFUDHMU","created_at":"2026-07-05T09:57:20.848872+00:00"},{"alias_kind":"pith_short_8","alias_value":"PFFUIBJ7","created_at":"2026-07-05T09:57:20.848872+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.17228","citing_title":"P3SL: Personalized Privacy-Preserving Split Learning on Heterogeneous Edge Devices","ref_index":44,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M","json":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M.json","graph_json":"https://pith.science/api/pith-number/PFFUIBJ7IMFUDHMUXRDKNEUT2M/graph.json","events_json":"https://pith.science/api/pith-number/PFFUIBJ7IMFUDHMUXRDKNEUT2M/events.json","paper":"https://pith.science/paper/PFFUIBJ7"},"agent_actions":{"view_html":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M","download_json":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M.json","view_paper":"https://pith.science/paper/PFFUIBJ7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.05285&json=true","fetch_graph":"https://pith.science/api/pith-number/PFFUIBJ7IMFUDHMUXRDKNEUT2M/graph.json","fetch_events":"https://pith.science/api/pith-number/PFFUIBJ7IMFUDHMUXRDKNEUT2M/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M/action/storage_attestation","attest_author":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M/action/author_attestation","sign_citation":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M/action/citation_signature","submit_replication":"https://pith.science/pith/PFFUIBJ7IMFUDHMUXRDKNEUT2M/action/replication_record"}},"created_at":"2026-07-05T09:57:20.848872+00:00","updated_at":"2026-07-05T09:57:20.848872+00:00"}