{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:MWDKT44RRR6G6M75L4CJBHLZFG","short_pith_number":"pith:MWDKT44R","schema_version":"1.0","canonical_sha256":"6586a9f3918c7c6f33fd5f04909d79299a0722fd10e78af9b7753dac781cc43f","source":{"kind":"arxiv","id":"2111.15521","version":3},"attestation_state":"computed","paper":{"title":"Node-Level Differentially Private Graph Neural Networks","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Abhradeep Guha Thakurta, Aditya Sinha, Ameya Daigavane, Gagan Madan, Gaurav Aggarwal, Prateek Jain","submitted_at":"2021-11-23T16:18:53Z","abstract_excerpt":"Graph Neural Networks (GNNs) are a popular technique for modelling graph-structured data and computing node-level representations via aggregation of information from the neighborhood of each node. However, this aggregation implies an increased risk of revealing sensitive information, as a node can participate in the inference for multiple nodes. This implies that standard privacy-preserving machine learning techniques, such as differentially private stochastic gradient descent (DP-SGD) - which are designed for situations where each data point participates in the inference for one point only - "},"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":"2111.15521","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2021-11-23T16:18:53Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"ec6a96884e3de4a36b2f7b3206eddee9bf07732e194bbd5a6bab8ad915d2794c","abstract_canon_sha256":"7243f6a839e8fcf63d96eb5e89c421f5890b57ec49ae98065d3dcae8cdac33fb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:51:53.748393Z","signature_b64":"9IQCBokTM2u7WXUM/IEwqsNiwINqt4PSbSTpuw2n7d3FzpW3/h0EEMfJtrd4pk1TisPT/68ZU2n/XZS84YksBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6586a9f3918c7c6f33fd5f04909d79299a0722fd10e78af9b7753dac781cc43f","last_reissued_at":"2026-07-05T04:51:53.747920Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:51:53.747920Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Node-Level Differentially Private Graph Neural Networks","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Abhradeep Guha Thakurta, Aditya Sinha, Ameya Daigavane, Gagan Madan, Gaurav Aggarwal, Prateek Jain","submitted_at":"2021-11-23T16:18:53Z","abstract_excerpt":"Graph Neural Networks (GNNs) are a popular technique for modelling graph-structured data and computing node-level representations via aggregation of information from the neighborhood of each node. However, this aggregation implies an increased risk of revealing sensitive information, as a node can participate in the inference for multiple nodes. This implies that standard privacy-preserving machine learning techniques, such as differentially private stochastic gradient descent (DP-SGD) - which are designed for situations where each data point participates in the inference for one point only - "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.15521","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/2111.15521/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":"2111.15521","created_at":"2026-07-05T04:51:53.747977+00:00"},{"alias_kind":"arxiv_version","alias_value":"2111.15521v3","created_at":"2026-07-05T04:51:53.747977+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.15521","created_at":"2026-07-05T04:51:53.747977+00:00"},{"alias_kind":"pith_short_12","alias_value":"MWDKT44RRR6G","created_at":"2026-07-05T04:51:53.747977+00:00"},{"alias_kind":"pith_short_16","alias_value":"MWDKT44RRR6G6M75","created_at":"2026-07-05T04:51:53.747977+00:00"},{"alias_kind":"pith_short_8","alias_value":"MWDKT44R","created_at":"2026-07-05T04:51:53.747977+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.26243","citing_title":"Provably Communication-Efficient and Privacy-Preserving Federated Graph Neural Networks","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01987","citing_title":"Misclassification Rate and Privacy-Utility Trade-offs in Graph Convolutional Networks via Subsampling Stability","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG","json":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG.json","graph_json":"https://pith.science/api/pith-number/MWDKT44RRR6G6M75L4CJBHLZFG/graph.json","events_json":"https://pith.science/api/pith-number/MWDKT44RRR6G6M75L4CJBHLZFG/events.json","paper":"https://pith.science/paper/MWDKT44R"},"agent_actions":{"view_html":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG","download_json":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG.json","view_paper":"https://pith.science/paper/MWDKT44R","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2111.15521&json=true","fetch_graph":"https://pith.science/api/pith-number/MWDKT44RRR6G6M75L4CJBHLZFG/graph.json","fetch_events":"https://pith.science/api/pith-number/MWDKT44RRR6G6M75L4CJBHLZFG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG/action/storage_attestation","attest_author":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG/action/author_attestation","sign_citation":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG/action/citation_signature","submit_replication":"https://pith.science/pith/MWDKT44RRR6G6M75L4CJBHLZFG/action/replication_record"}},"created_at":"2026-07-05T04:51:53.747977+00:00","updated_at":"2026-07-05T04:51:53.747977+00:00"}