{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LS7HDLRBAYOUEWYOOMLWLZVOGJ","short_pith_number":"pith:LS7HDLRB","schema_version":"1.0","canonical_sha256":"5cbe71ae21061d425b0e731765e6ae32494b346bc6befa8def8fbbd7f5a2226a","source":{"kind":"arxiv","id":"2508.08962","version":1},"attestation_state":"computed","paper":{"title":"Selection of Layers from Self-supervised Learning Models for Predicting Mean-Opinion-Score of Speech","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"eess.AS","authors_text":"Chandan K. A. Reddy, Christian Schuldt, Fredrik Cumlin, Saikat Chatterjee, Victor Ungureanu, Xinyu Liang","submitted_at":"2025-08-12T14:25:55Z","abstract_excerpt":"Self-supervised learning (SSL) models like Wav2Vec2, HuBERT, and WavLM have been widely used in speech processing. These transformer-based models consist of multiple layers, each capturing different levels of representation. While prior studies explored their layer-wise representations for efficiency and performance, speech quality assessment (SQA) models predominantly rely on last-layer features, leaving intermediate layers underexamined. In this work, we systematically evaluate different layers of multiple SSL models for predicting mean-opinion-score (MOS). Features from each layer are fed i"},"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":"2508.08962","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2025-08-12T14:25:55Z","cross_cats_sorted":["cs.SD"],"title_canon_sha256":"b7d98d57eff6d7243a3449bbd1309796dfe003e220859340735adb2197dab5de","abstract_canon_sha256":"f9e5e4c0a30229d4bd9f9cc7f2722ecfc36c020faed7ba16cb7097bf0cf554c4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:52:37.590389Z","signature_b64":"IYTItqC9B5NL+ww6w2JS2HFD3+LiqXWLEfMeJlLyTY+pBDcXsAPaOpGqyECty1Mf2av6UWm3ullzk3thpzqDDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5cbe71ae21061d425b0e731765e6ae32494b346bc6befa8def8fbbd7f5a2226a","last_reissued_at":"2026-07-05T11:52:37.589870Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:52:37.589870Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Selection of Layers from Self-supervised Learning Models for Predicting Mean-Opinion-Score of Speech","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"eess.AS","authors_text":"Chandan K. A. Reddy, Christian Schuldt, Fredrik Cumlin, Saikat Chatterjee, Victor Ungureanu, Xinyu Liang","submitted_at":"2025-08-12T14:25:55Z","abstract_excerpt":"Self-supervised learning (SSL) models like Wav2Vec2, HuBERT, and WavLM have been widely used in speech processing. These transformer-based models consist of multiple layers, each capturing different levels of representation. While prior studies explored their layer-wise representations for efficiency and performance, speech quality assessment (SQA) models predominantly rely on last-layer features, leaving intermediate layers underexamined. In this work, we systematically evaluate different layers of multiple SSL models for predicting mean-opinion-score (MOS). Features from each layer are fed i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.08962","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/2508.08962/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":"2508.08962","created_at":"2026-07-05T11:52:37.589930+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.08962v1","created_at":"2026-07-05T11:52:37.589930+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.08962","created_at":"2026-07-05T11:52:37.589930+00:00"},{"alias_kind":"pith_short_12","alias_value":"LS7HDLRBAYOU","created_at":"2026-07-05T11:52:37.589930+00:00"},{"alias_kind":"pith_short_16","alias_value":"LS7HDLRBAYOUEWYO","created_at":"2026-07-05T11:52:37.589930+00:00"},{"alias_kind":"pith_short_8","alias_value":"LS7HDLRB","created_at":"2026-07-05T11:52:37.589930+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26903","citing_title":"DNSMOS-C: Improving End-to-end Speech Quality Models via Contrastive Learning","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14231","citing_title":"AudioMosaic: Contrastive Masked Audio Representation Learning","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ","json":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ.json","graph_json":"https://pith.science/api/pith-number/LS7HDLRBAYOUEWYOOMLWLZVOGJ/graph.json","events_json":"https://pith.science/api/pith-number/LS7HDLRBAYOUEWYOOMLWLZVOGJ/events.json","paper":"https://pith.science/paper/LS7HDLRB"},"agent_actions":{"view_html":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ","download_json":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ.json","view_paper":"https://pith.science/paper/LS7HDLRB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.08962&json=true","fetch_graph":"https://pith.science/api/pith-number/LS7HDLRBAYOUEWYOOMLWLZVOGJ/graph.json","fetch_events":"https://pith.science/api/pith-number/LS7HDLRBAYOUEWYOOMLWLZVOGJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ/action/storage_attestation","attest_author":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ/action/author_attestation","sign_citation":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ/action/citation_signature","submit_replication":"https://pith.science/pith/LS7HDLRBAYOUEWYOOMLWLZVOGJ/action/replication_record"}},"created_at":"2026-07-05T11:52:37.589930+00:00","updated_at":"2026-07-05T11:52:37.589930+00:00"}