{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2JGWRGTJ6VLXYMU3V3ATK364P5","short_pith_number":"pith:2JGWRGTJ","schema_version":"1.0","canonical_sha256":"d24d689a69f5577c329baec1356fdc7f5270cdb13df615597f256f8fc8042d25","source":{"kind":"arxiv","id":"2508.15860","version":4},"attestation_state":"computed","paper":{"title":"Robust Residual Finite Scalar Quantization for Neural Compression","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV","eess.AS"],"primary_cat":"eess.IV","authors_text":"Baoxiang Li, Dongchuan Ran, Xiaoxu Zhu, Yiming Ren","submitted_at":"2025-08-20T15:18:59Z","abstract_excerpt":"Finite Scalar Quantization (FSQ) offers simplified training but suffers from residual magnitude decay in multi-stage settings, where subsequent stages receive exponentially weaker signals. We propose Robust Residual Finite Scalar Quantization (RFSQ), addressing this fundamental limitation through two novel conditioning strategies: learnable scaling factors and invertible layer normalization. Our experiments across audio and image modalities demonstrate RFSQ's effectiveness and generalizability. In audio reconstruction at 24 bits/frame, RFSQ-LayerNorm achieves 3.646 DNSMOS, a 3.6% improvement o"},"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":"2508.15860","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.IV","submitted_at":"2025-08-20T15:18:59Z","cross_cats_sorted":["cs.CV","eess.AS"],"title_canon_sha256":"3500bb5066df4aa6d44eb8458e053a230f79e2c9e8a3b778f3e05287f02edb3c","abstract_canon_sha256":"394d9a729f68d09905fb4464b0f8baa901ac0f670a03b6bdd0fc0bc790d8c118"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d24d689a69f5577c329baec1356fdc7f5270cdb13df615597f256f8fc8042d25","last_reissued_at":"2026-07-31T01:32:37.688389Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-31T01:32:37.688389Z"},"graph_snapshot":{"paper":{"title":"Robust Residual Finite Scalar Quantization for Neural Compression","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV","eess.AS"],"primary_cat":"eess.IV","authors_text":"Baoxiang Li, Dongchuan Ran, Xiaoxu Zhu, Yiming Ren","submitted_at":"2025-08-20T15:18:59Z","abstract_excerpt":"Finite Scalar Quantization (FSQ) offers simplified training but suffers from residual magnitude decay in multi-stage settings, where subsequent stages receive exponentially weaker signals. We propose Robust Residual Finite Scalar Quantization (RFSQ), addressing this fundamental limitation through two novel conditioning strategies: learnable scaling factors and invertible layer normalization. Our experiments across audio and image modalities demonstrate RFSQ's effectiveness and generalizability. In audio reconstruction at 24 bits/frame, RFSQ-LayerNorm achieves 3.646 DNSMOS, a 3.6% improvement o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.15860","kind":"arxiv","version":4},"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/2508.15860/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":"2508.15860","created_at":"2026-07-31T01:32:37.691440+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.15860v4","created_at":"2026-07-31T01:32:37.691440+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.15860","created_at":"2026-07-31T01:32:37.691440+00:00"},{"alias_kind":"pith_short_12","alias_value":"2JGWRGTJ6VLX","created_at":"2026-07-31T01:32:37.691440+00:00"},{"alias_kind":"pith_short_16","alias_value":"2JGWRGTJ6VLXYMU3","created_at":"2026-07-31T01:32:37.691440+00:00"},{"alias_kind":"pith_short_8","alias_value":"2JGWRGTJ","created_at":"2026-07-31T01:32:37.691440+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.23373","citing_title":"AffectCodec: Emotion-Preserving Neural Speech Codec with Block-Diagonal Residual FSQ","ref_index":32,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5","json":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5.json","graph_json":"https://pith.science/api/pith-number/2JGWRGTJ6VLXYMU3V3ATK364P5/graph.json","events_json":"https://pith.science/api/pith-number/2JGWRGTJ6VLXYMU3V3ATK364P5/events.json","paper":"https://pith.science/paper/2JGWRGTJ"},"agent_actions":{"view_html":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5","download_json":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5.json","view_paper":"https://pith.science/paper/2JGWRGTJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.15860&json=true","fetch_graph":"https://pith.science/api/pith-number/2JGWRGTJ6VLXYMU3V3ATK364P5/graph.json","fetch_events":"https://pith.science/api/pith-number/2JGWRGTJ6VLXYMU3V3ATK364P5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5/action/storage_attestation","attest_author":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5/action/author_attestation","sign_citation":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5/action/citation_signature","submit_replication":"https://pith.science/pith/2JGWRGTJ6VLXYMU3V3ATK364P5/action/replication_record"}},"created_at":"2026-07-31T01:32:37.691440+00:00","updated_at":"2026-07-31T01:32:37.691440+00:00"}