{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NNNMDLXU36FLBKPE2F2ZTLX3AK","short_pith_number":"pith:NNNMDLXU","schema_version":"1.0","canonical_sha256":"6b5ac1aef4df8ab0a9e4d17599aefb028d522f617bdf209a1b51ce9a6173bef3","source":{"kind":"arxiv","id":"2607.06392","version":1},"attestation_state":"computed","paper":{"title":"InsideSSL: Understanding Self-Supervised Speech Representations using a Model-Centric Perspective","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SD","authors_text":"Samir Sadok, Xavier Alameda-Pineda","submitted_at":"2026-07-07T15:25:47Z","abstract_excerpt":"Self-supervised learning (SSL) models, such as Wav2Vec2, HuBERT, and WavLM, have become foundational across a wide range of speech and audio tasks. Despite their success, understanding their internal layer-wise dynamics remains an ongoing challenge. To address this, we propose a two-part model-centric framework called InsideSSL. First, we establish a task-agnostic analysis from three intrinsic per-layer perspectives: compression (entropy), geometry (curvature), and robustness to perturbations. We show that varying training objectives induce distinct regimes of acoustic compression and manifold"},"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":"2607.06392","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.SD","submitted_at":"2026-07-07T15:25:47Z","cross_cats_sorted":[],"title_canon_sha256":"b3bc626d96fc6f910d0fea4bf971137981c6d6a48e66680fb2ff14c4fa169bc4","abstract_canon_sha256":"35c5b589f5b7c7ef700305f2d460fa4d57c64b73786a9450bbd2a463c048cb6b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-08T01:19:24.054760Z","signature_b64":"Sxf5ajO6nM92z5pGPltuQDJdJprUFndEzd9FyYdS58064/YhrcSu9at9pMqOxENqzV3LzxsYVrA5QrxGgxB/Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6b5ac1aef4df8ab0a9e4d17599aefb028d522f617bdf209a1b51ce9a6173bef3","last_reissued_at":"2026-07-08T01:19:24.054295Z","signature_status":"signed_v1","first_computed_at":"2026-07-08T01:19:24.054295Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"InsideSSL: Understanding Self-Supervised Speech Representations using a Model-Centric Perspective","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SD","authors_text":"Samir Sadok, Xavier Alameda-Pineda","submitted_at":"2026-07-07T15:25:47Z","abstract_excerpt":"Self-supervised learning (SSL) models, such as Wav2Vec2, HuBERT, and WavLM, have become foundational across a wide range of speech and audio tasks. Despite their success, understanding their internal layer-wise dynamics remains an ongoing challenge. To address this, we propose a two-part model-centric framework called InsideSSL. First, we establish a task-agnostic analysis from three intrinsic per-layer perspectives: compression (entropy), geometry (curvature), and robustness to perturbations. We show that varying training objectives induce distinct regimes of acoustic compression and manifold"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.06392","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/2607.06392/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":"2607.06392","created_at":"2026-07-08T01:19:24.054366+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.06392v1","created_at":"2026-07-08T01:19:24.054366+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.06392","created_at":"2026-07-08T01:19:24.054366+00:00"},{"alias_kind":"pith_short_12","alias_value":"NNNMDLXU36FL","created_at":"2026-07-08T01:19:24.054366+00:00"},{"alias_kind":"pith_short_16","alias_value":"NNNMDLXU36FLBKPE","created_at":"2026-07-08T01:19:24.054366+00:00"},{"alias_kind":"pith_short_8","alias_value":"NNNMDLXU","created_at":"2026-07-08T01:19:24.054366+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06392","citing_title":"InsideSSL: Understanding Self-Supervised Speech Representations using a Model-Centric Perspective","ref_index":2,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK","json":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK.json","graph_json":"https://pith.science/api/pith-number/NNNMDLXU36FLBKPE2F2ZTLX3AK/graph.json","events_json":"https://pith.science/api/pith-number/NNNMDLXU36FLBKPE2F2ZTLX3AK/events.json","paper":"https://pith.science/paper/NNNMDLXU"},"agent_actions":{"view_html":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK","download_json":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK.json","view_paper":"https://pith.science/paper/NNNMDLXU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.06392&json=true","fetch_graph":"https://pith.science/api/pith-number/NNNMDLXU36FLBKPE2F2ZTLX3AK/graph.json","fetch_events":"https://pith.science/api/pith-number/NNNMDLXU36FLBKPE2F2ZTLX3AK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK/action/storage_attestation","attest_author":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK/action/author_attestation","sign_citation":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK/action/citation_signature","submit_replication":"https://pith.science/pith/NNNMDLXU36FLBKPE2F2ZTLX3AK/action/replication_record"}},"created_at":"2026-07-08T01:19:24.054366+00:00","updated_at":"2026-07-08T01:19:24.054366+00:00"}