{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:WW5XYOQVULYG4GQXJGCSX44TWC","short_pith_number":"pith:WW5XYOQV","schema_version":"1.0","canonical_sha256":"b5bb7c3a15a2f06e1a1749852bf393b0bfb69975d8bd069035ac76bb6196a918","source":{"kind":"arxiv","id":"2503.12008","version":1},"attestation_state":"computed","paper":{"title":"Winning the MIDST Challenge: New Membership Inference Attacks on Diffusion Models for Tabular Data Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Steven Wu, Terrance Liu, Xiaoyu Wu, Yifei Pang","submitted_at":"2025-03-15T06:13:27Z","abstract_excerpt":"Tabular data synthesis using diffusion models has gained significant attention for its potential to balance data utility and privacy. However, existing privacy evaluations often rely on heuristic metrics or weak membership inference attacks (MIA), leaving privacy risks inadequately assessed. In this work, we conduct a rigorous MIA study on diffusion-based tabular synthesis, revealing that state-of-the-art attacks designed for image models fail in this setting. We identify noise initialization as a key factor influencing attack efficacy and propose a machine-learning-driven approach that levera"},"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":"2503.12008","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-15T06:13:27Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"686109d0de4cf0f6b140dd58901db28b0cc88e2cb439b09498cf4cc159499c28","abstract_canon_sha256":"511672a51b6b3a3be782a27d1a19d9f7857810e59a579b091ceae34fcf08bd9f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:32:42.647296Z","signature_b64":"4uT9dziU2gff3MToI7MD3quqe27G58h7eYSA7ctWzsOKZXCTr6JgJ7LjqRCKB2keeRRP+hj3NJoUmxVUFzazCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b5bb7c3a15a2f06e1a1749852bf393b0bfb69975d8bd069035ac76bb6196a918","last_reissued_at":"2026-07-05T10:32:42.646482Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:32:42.646482Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Winning the MIDST Challenge: New Membership Inference Attacks on Diffusion Models for Tabular Data Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Steven Wu, Terrance Liu, Xiaoyu Wu, Yifei Pang","submitted_at":"2025-03-15T06:13:27Z","abstract_excerpt":"Tabular data synthesis using diffusion models has gained significant attention for its potential to balance data utility and privacy. However, existing privacy evaluations often rely on heuristic metrics or weak membership inference attacks (MIA), leaving privacy risks inadequately assessed. In this work, we conduct a rigorous MIA study on diffusion-based tabular synthesis, revealing that state-of-the-art attacks designed for image models fail in this setting. We identify noise initialization as a key factor influencing attack efficacy and propose a machine-learning-driven approach that levera"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.12008","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/2503.12008/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":"2503.12008","created_at":"2026-07-05T10:32:42.646569+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.12008v1","created_at":"2026-07-05T10:32:42.646569+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.12008","created_at":"2026-07-05T10:32:42.646569+00:00"},{"alias_kind":"pith_short_12","alias_value":"WW5XYOQVULYG","created_at":"2026-07-05T10:32:42.646569+00:00"},{"alias_kind":"pith_short_16","alias_value":"WW5XYOQVULYG4GQX","created_at":"2026-07-05T10:32:42.646569+00:00"},{"alias_kind":"pith_short_8","alias_value":"WW5XYOQV","created_at":"2026-07-05T10:32:42.646569+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.11527","citing_title":"FERMI: Exploiting Relations for Membership Inference Against Tabular Diffusion Models","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19653","citing_title":"A Dual Perspective on Synthetic Trajectory Generators: Utility Framework and Privacy Vulnerabilities","ref_index":122,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06835","citing_title":"On Privacy Leakage in Tabular Diffusion Models: Influential Factors, Attacker Knowledge, and Metrics","ref_index":61,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC","json":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC.json","graph_json":"https://pith.science/api/pith-number/WW5XYOQVULYG4GQXJGCSX44TWC/graph.json","events_json":"https://pith.science/api/pith-number/WW5XYOQVULYG4GQXJGCSX44TWC/events.json","paper":"https://pith.science/paper/WW5XYOQV"},"agent_actions":{"view_html":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC","download_json":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC.json","view_paper":"https://pith.science/paper/WW5XYOQV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.12008&json=true","fetch_graph":"https://pith.science/api/pith-number/WW5XYOQVULYG4GQXJGCSX44TWC/graph.json","fetch_events":"https://pith.science/api/pith-number/WW5XYOQVULYG4GQXJGCSX44TWC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC/action/storage_attestation","attest_author":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC/action/author_attestation","sign_citation":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC/action/citation_signature","submit_replication":"https://pith.science/pith/WW5XYOQVULYG4GQXJGCSX44TWC/action/replication_record"}},"created_at":"2026-07-05T10:32:42.646569+00:00","updated_at":"2026-07-05T10:32:42.646569+00:00"}