{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:MI4EFQ7APG72CKVGZWE2MQLCQG","short_pith_number":"pith:MI4EFQ7A","schema_version":"1.0","canonical_sha256":"623842c3e079bfa12aa6cd89a6416281a1b7bc29e132e90804af7ff2f538520e","source":{"kind":"arxiv","id":"2506.22362","version":1},"attestation_state":"computed","paper":{"title":"DiffSoundStream: Efficient Speech Tokenization via Diffusion Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"eess.AS","authors_text":"Alessio Centazzo, Craig Dooley, George Sung, Ramanan Rajeswaran, Shao-fu Shih, Yang Yang, Yunpeng Li","submitted_at":"2025-06-27T16:23:07Z","abstract_excerpt":"Token-based language modeling is a prominent approach for speech generation, where tokens are obtained by quantizing features from self-supervised learning (SSL) models and extracting codes from neural speech codecs, generally referred to as semantic tokens and acoustic tokens. These tokens are often modeled autoregressively, with the inference speed being constrained by the token rate. In this work, we propose DiffSoundStream, a solution that improves the efficiency of speech tokenization in non-streaming scenarios through two techniques: (1) conditioning the neural codec on semantic tokens t"},"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":"2506.22362","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2025-06-27T16:23:07Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"680649cb74f150394d47b8f2e02eb415c39bdd1cc7eee2278cc05af799fadf62","abstract_canon_sha256":"7dee1febc4e9d8f181ccf3d9b6e526ba6e884ec73b16d413b82bcad20a9f2537"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:28:26.045543Z","signature_b64":"yw4ldSBshgYCAtuDOXJu8zM27FpjNxSp6b2mUevFCuEd7ZTeVaxELHELBz35kkLAZLcA00UQG1FyK6z2k+OtCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"623842c3e079bfa12aa6cd89a6416281a1b7bc29e132e90804af7ff2f538520e","last_reissued_at":"2026-07-05T11:28:26.045035Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:28:26.045035Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DiffSoundStream: Efficient Speech Tokenization via Diffusion Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"eess.AS","authors_text":"Alessio Centazzo, Craig Dooley, George Sung, Ramanan Rajeswaran, Shao-fu Shih, Yang Yang, Yunpeng Li","submitted_at":"2025-06-27T16:23:07Z","abstract_excerpt":"Token-based language modeling is a prominent approach for speech generation, where tokens are obtained by quantizing features from self-supervised learning (SSL) models and extracting codes from neural speech codecs, generally referred to as semantic tokens and acoustic tokens. These tokens are often modeled autoregressively, with the inference speed being constrained by the token rate. In this work, we propose DiffSoundStream, a solution that improves the efficiency of speech tokenization in non-streaming scenarios through two techniques: (1) conditioning the neural codec on semantic tokens t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.22362","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/2506.22362/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":"2506.22362","created_at":"2026-07-05T11:28:26.045092+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.22362v1","created_at":"2026-07-05T11:28:26.045092+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.22362","created_at":"2026-07-05T11:28:26.045092+00:00"},{"alias_kind":"pith_short_12","alias_value":"MI4EFQ7APG72","created_at":"2026-07-05T11:28:26.045092+00:00"},{"alias_kind":"pith_short_16","alias_value":"MI4EFQ7APG72CKVG","created_at":"2026-07-05T11:28:26.045092+00:00"},{"alias_kind":"pith_short_8","alias_value":"MI4EFQ7A","created_at":"2026-07-05T11:28:26.045092+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.22362","citing_title":"DiffSoundStream: Efficient Speech Tokenization via Diffusion Decoding","ref_index":1,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG","json":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG.json","graph_json":"https://pith.science/api/pith-number/MI4EFQ7APG72CKVGZWE2MQLCQG/graph.json","events_json":"https://pith.science/api/pith-number/MI4EFQ7APG72CKVGZWE2MQLCQG/events.json","paper":"https://pith.science/paper/MI4EFQ7A"},"agent_actions":{"view_html":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG","download_json":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG.json","view_paper":"https://pith.science/paper/MI4EFQ7A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.22362&json=true","fetch_graph":"https://pith.science/api/pith-number/MI4EFQ7APG72CKVGZWE2MQLCQG/graph.json","fetch_events":"https://pith.science/api/pith-number/MI4EFQ7APG72CKVGZWE2MQLCQG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG/action/storage_attestation","attest_author":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG/action/author_attestation","sign_citation":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG/action/citation_signature","submit_replication":"https://pith.science/pith/MI4EFQ7APG72CKVGZWE2MQLCQG/action/replication_record"}},"created_at":"2026-07-05T11:28:26.045092+00:00","updated_at":"2026-07-05T11:28:26.045092+00:00"}