{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UMFYPHPDLU5VXCUF7QAGF2VZ5S","short_pith_number":"pith:UMFYPHPD","schema_version":"1.0","canonical_sha256":"a30b879de35d3b5b8a85fc0062eab9ec90fdfedd60b557c2d2363e27c7ab3d14","source":{"kind":"arxiv","id":"2608.12082","version":1},"attestation_state":"computed","paper":{"title":"Rethinking Language Model-Based Generative Speech Enhancement in the Latent Space of a Neural Audio Codec","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"eess.AS","authors_text":"Tim Fingscheidt, Yihui Fu, Zhengyang Li","submitted_at":"2026-08-12T14:04:37Z","abstract_excerpt":"Language model (LM)-based speech enhancement (SE) has recently emerged rapidly using latent space features of neural audio codecs (NACs). In this paper, first, we present a unified framework covering six popular LM-based generative SE modeling paradigms based on discrete/continuous latent NAC features: discrete or continuous autoregressive (D/CAR) SE, discrete or continuous non-autoregressive (D/CNAR) SE, discrete diffusion (DDiff) SE, and continuous flow matching (CFM) SE. Second, we are the first to compare their performance in a unified experimental setup and synopsis with diverse intrusive"},"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":"2608.12082","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2026-08-12T14:04:37Z","cross_cats_sorted":[],"title_canon_sha256":"1e647e4a43b4f87e2ea85fc1d5b212a48fc95c823fa3f1908f96496ab25ce997","abstract_canon_sha256":"8d2d0c160783d86232881ed497ef766f5e1a38e96c761a5f130736f852d04da4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-13T01:29:26.962537Z","signature_b64":"4IBUJ1GnevqmMSpdJVSwVdQ9lD5VaTRmvJRN0kicD+6fJZ91fXRVnmPIBiazGHz4qcIH8HMmwb3AgPCpDub+Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a30b879de35d3b5b8a85fc0062eab9ec90fdfedd60b557c2d2363e27c7ab3d14","last_reissued_at":"2026-08-13T01:29:26.960229Z","signature_status":"signed_v1","first_computed_at":"2026-08-13T01:29:26.960229Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethinking Language Model-Based Generative Speech Enhancement in the Latent Space of a Neural Audio Codec","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"eess.AS","authors_text":"Tim Fingscheidt, Yihui Fu, Zhengyang Li","submitted_at":"2026-08-12T14:04:37Z","abstract_excerpt":"Language model (LM)-based speech enhancement (SE) has recently emerged rapidly using latent space features of neural audio codecs (NACs). In this paper, first, we present a unified framework covering six popular LM-based generative SE modeling paradigms based on discrete/continuous latent NAC features: discrete or continuous autoregressive (D/CAR) SE, discrete or continuous non-autoregressive (D/CNAR) SE, discrete diffusion (DDiff) SE, and continuous flow matching (CFM) SE. Second, we are the first to compare their performance in a unified experimental setup and synopsis with diverse intrusive"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.12082","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/2608.12082/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":"2608.12082","created_at":"2026-08-13T01:29:26.961478+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.12082v1","created_at":"2026-08-13T01:29:26.961478+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.12082","created_at":"2026-08-13T01:29:26.961478+00:00"},{"alias_kind":"pith_short_12","alias_value":"UMFYPHPDLU5V","created_at":"2026-08-13T01:29:26.961478+00:00"},{"alias_kind":"pith_short_16","alias_value":"UMFYPHPDLU5VXCUF","created_at":"2026-08-13T01:29:26.961478+00:00"},{"alias_kind":"pith_short_8","alias_value":"UMFYPHPD","created_at":"2026-08-13T01:29:26.961478+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.12082","citing_title":"Rethinking Language Model-Based Generative Speech Enhancement in the Latent Space of a Neural Audio Codec","ref_index":2,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S","json":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S.json","graph_json":"https://pith.science/api/pith-number/UMFYPHPDLU5VXCUF7QAGF2VZ5S/graph.json","events_json":"https://pith.science/api/pith-number/UMFYPHPDLU5VXCUF7QAGF2VZ5S/events.json","paper":"https://pith.science/paper/UMFYPHPD"},"agent_actions":{"view_html":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S","download_json":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S.json","view_paper":"https://pith.science/paper/UMFYPHPD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.12082&json=true","fetch_graph":"https://pith.science/api/pith-number/UMFYPHPDLU5VXCUF7QAGF2VZ5S/graph.json","fetch_events":"https://pith.science/api/pith-number/UMFYPHPDLU5VXCUF7QAGF2VZ5S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S/action/storage_attestation","attest_author":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S/action/author_attestation","sign_citation":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S/action/citation_signature","submit_replication":"https://pith.science/pith/UMFYPHPDLU5VXCUF7QAGF2VZ5S/action/replication_record"}},"created_at":"2026-08-13T01:29:26.961478+00:00","updated_at":"2026-08-13T01:29:26.961478+00:00"}