{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:QSSPSIEGDFLYLZFTWO2AGROTQJ","short_pith_number":"pith:QSSPSIEG","schema_version":"1.0","canonical_sha256":"84a4f92086195785e4b3b3b40345d3824d89ee8a91b9d529c5115cb620687b63","source":{"kind":"arxiv","id":"2606.19398","version":1},"attestation_state":"computed","paper":{"title":"S-JEPA : Soft Clustering Anchors for Self-Supervised Speech Representation Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["eess.AS","eess.SP"],"primary_cat":"cs.SD","authors_text":"Aaron Elkins, Adrian Kieback, Aman Chadha, Georgios Ioannides, Judah Goldfeder, Linsey Pang, Ravid Shwartz-Ziv, Yann LeCun","submitted_at":"2026-06-17T08:39:09Z","abstract_excerpt":"Self-supervised speech encoders are predominantly trained by predicting discrete hard cluster IDs at masked positions, a recipe that collapses acoustic ambiguity at category boundaries and requires interrupting training to re-cluster the entire corpus between iterations. We introduce S-JEPA, a JEPA-style encoder-predictor pair trained to match the soft posteriors of a Gaussian Mixture Model at masked positions via KL divergence. Training runs as one continuous optimization trajectory in two phases: a fixed GMM over MFCC features, then an online GMM over encoder features, with the input layer s"},"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":"2606.19398","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SD","submitted_at":"2026-06-17T08:39:09Z","cross_cats_sorted":["eess.AS","eess.SP"],"title_canon_sha256":"c647fa3102200075f0bb2a7414fd29a790f3b1bba0dd6510cef06aacdae405ef","abstract_canon_sha256":"388678ef1108ef5658009a4780ea84dc86acde6e905df5c1b1a50f449df1b7f1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:12:24.533354Z","signature_b64":"SbyNDuZApPIshQfzJ5Tn0EPqlxFIHAEPdAufmhn0C1l9G6rZAC/VyQSei12niQ92GRy4ymDlAiUDyo9TndHHAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"84a4f92086195785e4b3b3b40345d3824d89ee8a91b9d529c5115cb620687b63","last_reissued_at":"2026-06-19T16:12:24.532994Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:12:24.532994Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"S-JEPA : Soft Clustering Anchors for Self-Supervised Speech Representation Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["eess.AS","eess.SP"],"primary_cat":"cs.SD","authors_text":"Aaron Elkins, Adrian Kieback, Aman Chadha, Georgios Ioannides, Judah Goldfeder, Linsey Pang, Ravid Shwartz-Ziv, Yann LeCun","submitted_at":"2026-06-17T08:39:09Z","abstract_excerpt":"Self-supervised speech encoders are predominantly trained by predicting discrete hard cluster IDs at masked positions, a recipe that collapses acoustic ambiguity at category boundaries and requires interrupting training to re-cluster the entire corpus between iterations. We introduce S-JEPA, a JEPA-style encoder-predictor pair trained to match the soft posteriors of a Gaussian Mixture Model at masked positions via KL divergence. Training runs as one continuous optimization trajectory in two phases: a fixed GMM over MFCC features, then an online GMM over encoder features, with the input layer s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.19398","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/2606.19398/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":"2606.19398","created_at":"2026-06-19T16:12:24.533055+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.19398v1","created_at":"2026-06-19T16:12:24.533055+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.19398","created_at":"2026-06-19T16:12:24.533055+00:00"},{"alias_kind":"pith_short_12","alias_value":"QSSPSIEGDFLY","created_at":"2026-06-19T16:12:24.533055+00:00"},{"alias_kind":"pith_short_16","alias_value":"QSSPSIEGDFLYLZFT","created_at":"2026-06-19T16:12:24.533055+00:00"},{"alias_kind":"pith_short_8","alias_value":"QSSPSIEG","created_at":"2026-06-19T16:12:24.533055+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/QSSPSIEGDFLYLZFTWO2AGROTQJ","json":"https://pith.science/pith/QSSPSIEGDFLYLZFTWO2AGROTQJ.json","graph_json":"https://pith.science/api/pith-number/QSSPSIEGDFLYLZFTWO2AGROTQJ/graph.json","events_json":"https://pith.science/api/pith-number/QSSPSIEGDFLYLZFTWO2AGROTQJ/events.json","paper":"https://pith.science/paper/QSSPSIEG"},"agent_actions":{"view_html":"https://pith.science/pith/QSSPSIEGDFLYLZFTWO2AGROTQJ","download_json":"https://pith.science/pith/QSSPSIEGDFLYLZFTWO2AGROTQJ.json","view_paper":"https://pith.science/paper/QSSPSIEG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.19398&json=true","fetch_graph":"https://pith.science/api/pith-number/QSSPSIEGDFLYLZFTWO2AGROTQJ/graph.json","fetch_events":"https://pith.science/api/pith-number/QSSPSIEGDFLYLZFTWO2AGROTQJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QSSPSIEGDFLYLZFTWO2AGROTQJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QSSPSIEGDFLYLZFTWO2AGROTQJ/action/storage_attestation","attest_author":"https://pith.science/pith/QSSPSIEGDFLYLZFTWO2AGROTQJ/action/author_attestation","sign_citation":"https://pith.science/pith/QSSPSIEGDFLYLZFTWO2AGROTQJ/action/citation_signature","submit_replication":"https://pith.science/pith/QSSPSIEGDFLYLZFTWO2AGROTQJ/action/replication_record"}},"created_at":"2026-06-19T16:12:24.533055+00:00","updated_at":"2026-06-19T16:12:24.533055+00:00"}