{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3AKKHCLRIJYERUX5YBX2LZR3Y4","short_pith_number":"pith:3AKKHCLR","schema_version":"1.0","canonical_sha256":"d814a38971427048d2fdc06fa5e63bc71238d71068088e1c4601abc62c878136","source":{"kind":"arxiv","id":"2606.10912","version":1},"attestation_state":"computed","paper":{"title":"What Do Deepfake Speech Detectors Actually Hear?","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CR","cs.LG"],"primary_cat":"cs.SD","authors_text":"Anton Firc, Jakub Re\\v{s}, Kamil Malinka, Martin Pere\\v{s}\\'ini, Veronika Jirmusov\\'a, Vojt\\v{e}ch Stan\\v{e}k","submitted_at":"2026-06-09T14:21:45Z","abstract_excerpt":"Deepfake speech detectors often output a single score without explaining why an audio sample is flagged, where in the signal the evidence lies, or what cues drive the decision. We propose an audio-native explainability pipeline using Integrated Gradients on time-aligned self-supervised representations to localize decision evidence over time. We apply the proposed method to three WavLM-based detectors (AASIST, CA-MHFA, SLS) on ASVspoof 5 and manually annotate the highest-attribution regions to provide a semantic meaning of the most important cues. Despite similar performance, the detectors rely"},"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":"2606.10912","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.SD","submitted_at":"2026-06-09T14:21:45Z","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"title_canon_sha256":"923235060c7cb1038e7ec923116a4bbf9727c4504dc3b56cf354e862e51b073a","abstract_canon_sha256":"6d464e2d544f6b0cb2cd132a24e27b4069a7878d63e9dd2c84dc472c4f45d8c4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:10:47.737580Z","signature_b64":"YZih55Bm1o5STs3WFuIekho/7g0UAIBzMj0oLuRtHaCAzzdP/WmpMbBC9yOq+pSa7M+tFr/1TVmisnjZa3wBAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d814a38971427048d2fdc06fa5e63bc71238d71068088e1c4601abc62c878136","last_reissued_at":"2026-06-10T01:10:47.736649Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:10:47.736649Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What Do Deepfake Speech Detectors Actually Hear?","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CR","cs.LG"],"primary_cat":"cs.SD","authors_text":"Anton Firc, Jakub Re\\v{s}, Kamil Malinka, Martin Pere\\v{s}\\'ini, Veronika Jirmusov\\'a, Vojt\\v{e}ch Stan\\v{e}k","submitted_at":"2026-06-09T14:21:45Z","abstract_excerpt":"Deepfake speech detectors often output a single score without explaining why an audio sample is flagged, where in the signal the evidence lies, or what cues drive the decision. We propose an audio-native explainability pipeline using Integrated Gradients on time-aligned self-supervised representations to localize decision evidence over time. We apply the proposed method to three WavLM-based detectors (AASIST, CA-MHFA, SLS) on ASVspoof 5 and manually annotate the highest-attribution regions to provide a semantic meaning of the most important cues. Despite similar performance, the detectors rely"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10912","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/2606.10912/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":"2606.10912","created_at":"2026-06-10T01:10:47.736808+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.10912v1","created_at":"2026-06-10T01:10:47.736808+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10912","created_at":"2026-06-10T01:10:47.736808+00:00"},{"alias_kind":"pith_short_12","alias_value":"3AKKHCLRIJYE","created_at":"2026-06-10T01:10:47.736808+00:00"},{"alias_kind":"pith_short_16","alias_value":"3AKKHCLRIJYERUX5","created_at":"2026-06-10T01:10:47.736808+00:00"},{"alias_kind":"pith_short_8","alias_value":"3AKKHCLR","created_at":"2026-06-10T01:10:47.736808+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4","json":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4.json","graph_json":"https://pith.science/api/pith-number/3AKKHCLRIJYERUX5YBX2LZR3Y4/graph.json","events_json":"https://pith.science/api/pith-number/3AKKHCLRIJYERUX5YBX2LZR3Y4/events.json","paper":"https://pith.science/paper/3AKKHCLR"},"agent_actions":{"view_html":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4","download_json":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4.json","view_paper":"https://pith.science/paper/3AKKHCLR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.10912&json=true","fetch_graph":"https://pith.science/api/pith-number/3AKKHCLRIJYERUX5YBX2LZR3Y4/graph.json","fetch_events":"https://pith.science/api/pith-number/3AKKHCLRIJYERUX5YBX2LZR3Y4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4/action/storage_attestation","attest_author":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4/action/author_attestation","sign_citation":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4/action/citation_signature","submit_replication":"https://pith.science/pith/3AKKHCLRIJYERUX5YBX2LZR3Y4/action/replication_record"}},"created_at":"2026-06-10T01:10:47.736808+00:00","updated_at":"2026-06-10T01:10:47.736808+00:00"}