{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:H2GL4JYOCSS6M7HYQZY7PDLSRP","short_pith_number":"pith:H2GL4JYO","schema_version":"1.0","canonical_sha256":"3e8cbe270e14a5e67cf88671f78d728be6307758cc4e8c01ee4dbeb328f70ae7","source":{"kind":"arxiv","id":"2410.18371","version":2},"attestation_state":"computed","paper":{"title":"Gibberish is All You Need for Membership Inference Detection in Contrastive Language-Audio Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","eess.AS"],"primary_cat":"cs.SD","authors_text":"Ruoxi Cheng, Shitong Shao, Shuirong Cao, Yizhong Ding, Zhiqiang Wang","submitted_at":"2024-10-24T02:26:57Z","abstract_excerpt":"Audio can disclose PII, particularly when combined with related text data. Therefore, it is essential to develop tools to detect privacy leakage in Contrastive Language-Audio Pretraining(CLAP). Existing MIAs need audio as input, risking exposure of voiceprint and requiring costly shadow models. We first propose PRMID, a membership inference detector based probability ranking given by CLAP, which does not require training shadow models but still requires both audio and text of the individual as input. To address these limitations, we then propose USMID, a textual unimodal speaker-level membersh"},"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":"2410.18371","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SD","submitted_at":"2024-10-24T02:26:57Z","cross_cats_sorted":["cs.AI","eess.AS"],"title_canon_sha256":"62d2410c8ac2b3fb2aebb34942c14b752394ff2ce6ebc10c3789c9c475e130fb","abstract_canon_sha256":"86f967bfcdfbd976c5bf83c4480685336acd7dbf5647587dc70c5f56b1892765"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:30:22.817560Z","signature_b64":"TZzpKvFJHIHnXK5pL2JhkKvmFU9XUptDlWpS/WBzUV9lbhDFBe1NeWky7j7vdLJwQCsxQTCuwfUqui8cfyoyDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3e8cbe270e14a5e67cf88671f78d728be6307758cc4e8c01ee4dbeb328f70ae7","last_reissued_at":"2026-07-05T09:30:22.817089Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:30:22.817089Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gibberish is All You Need for Membership Inference Detection in Contrastive Language-Audio Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","eess.AS"],"primary_cat":"cs.SD","authors_text":"Ruoxi Cheng, Shitong Shao, Shuirong Cao, Yizhong Ding, Zhiqiang Wang","submitted_at":"2024-10-24T02:26:57Z","abstract_excerpt":"Audio can disclose PII, particularly when combined with related text data. Therefore, it is essential to develop tools to detect privacy leakage in Contrastive Language-Audio Pretraining(CLAP). Existing MIAs need audio as input, risking exposure of voiceprint and requiring costly shadow models. We first propose PRMID, a membership inference detector based probability ranking given by CLAP, which does not require training shadow models but still requires both audio and text of the individual as input. To address these limitations, we then propose USMID, a textual unimodal speaker-level membersh"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.18371","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/2410.18371/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":"2410.18371","created_at":"2026-07-05T09:30:22.817149+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.18371v2","created_at":"2026-07-05T09:30:22.817149+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.18371","created_at":"2026-07-05T09:30:22.817149+00:00"},{"alias_kind":"pith_short_12","alias_value":"H2GL4JYOCSS6","created_at":"2026-07-05T09:30:22.817149+00:00"},{"alias_kind":"pith_short_16","alias_value":"H2GL4JYOCSS6M7HY","created_at":"2026-07-05T09:30:22.817149+00:00"},{"alias_kind":"pith_short_8","alias_value":"H2GL4JYO","created_at":"2026-07-05T09:30:22.817149+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.24252","citing_title":"A Reward-driven Automated Webshell Malicious-code Generator for Red-teaming","ref_index":4,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP","json":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP.json","graph_json":"https://pith.science/api/pith-number/H2GL4JYOCSS6M7HYQZY7PDLSRP/graph.json","events_json":"https://pith.science/api/pith-number/H2GL4JYOCSS6M7HYQZY7PDLSRP/events.json","paper":"https://pith.science/paper/H2GL4JYO"},"agent_actions":{"view_html":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP","download_json":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP.json","view_paper":"https://pith.science/paper/H2GL4JYO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.18371&json=true","fetch_graph":"https://pith.science/api/pith-number/H2GL4JYOCSS6M7HYQZY7PDLSRP/graph.json","fetch_events":"https://pith.science/api/pith-number/H2GL4JYOCSS6M7HYQZY7PDLSRP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP/action/storage_attestation","attest_author":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP/action/author_attestation","sign_citation":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP/action/citation_signature","submit_replication":"https://pith.science/pith/H2GL4JYOCSS6M7HYQZY7PDLSRP/action/replication_record"}},"created_at":"2026-07-05T09:30:22.817149+00:00","updated_at":"2026-07-05T09:30:22.817149+00:00"}