{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:YTJKYBUESP7J6DKDYXGPXLEWWB","short_pith_number":"pith:YTJKYBUE","schema_version":"1.0","canonical_sha256":"c4d2ac068493fe9f0d43c5ccfbac96b06bf09c517a35bba005a7763ce1fa4c76","source":{"kind":"arxiv","id":"2310.11230","version":4},"attestation_state":"computed","paper":{"title":"Zipformer: A faster and better encoder for automatic speech recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Daniel Povey, Fangjun Kuang, Liyong Guo, Long Lin, Wei Kang, Xiaoyu Yang, Yifan Yang, Zengrui Jin, Zengwei Yao","submitted_at":"2023-10-17T13:01:10Z","abstract_excerpt":"The Conformer has become the most popular encoder model for automatic speech recognition (ASR). It adds convolution modules to a transformer to learn both local and global dependencies. In this work we describe a faster, more memory-efficient, and better-performing transformer, called Zipformer. Modeling changes include: 1) a U-Net-like encoder structure where middle stacks operate at lower frame rates; 2) reorganized block structure with more modules, within which we re-use attention weights for efficiency; 3) a modified form of LayerNorm called BiasNorm allows us to retain some length inform"},"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":"2310.11230","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2023-10-17T13:01:10Z","cross_cats_sorted":["cs.LG","cs.SD"],"title_canon_sha256":"2e7202af5bca6dbe86d0bb5b43ab016b6eea1de722fe33fc86d1aab4663f7724","abstract_canon_sha256":"80c213f5294b3bf6790928db054b8917b4d0590c1bf3464fd6cc50d2fd725e0b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:06:11.734052Z","signature_b64":"UH0IF3BckENgRTFjmR2gNqn4EToJxfGhmKJ8fOYg6r7rkiG4Byodcma/Gpyfm2rAmKskVhTTN+y0DQlPy5VVAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c4d2ac068493fe9f0d43c5ccfbac96b06bf09c517a35bba005a7763ce1fa4c76","last_reissued_at":"2026-07-05T08:06:11.733408Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:06:11.733408Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Zipformer: A faster and better encoder for automatic speech recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Daniel Povey, Fangjun Kuang, Liyong Guo, Long Lin, Wei Kang, Xiaoyu Yang, Yifan Yang, Zengrui Jin, Zengwei Yao","submitted_at":"2023-10-17T13:01:10Z","abstract_excerpt":"The Conformer has become the most popular encoder model for automatic speech recognition (ASR). It adds convolution modules to a transformer to learn both local and global dependencies. In this work we describe a faster, more memory-efficient, and better-performing transformer, called Zipformer. Modeling changes include: 1) a U-Net-like encoder structure where middle stacks operate at lower frame rates; 2) reorganized block structure with more modules, within which we re-use attention weights for efficiency; 3) a modified form of LayerNorm called BiasNorm allows us to retain some length inform"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.11230","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/2310.11230/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":"2310.11230","created_at":"2026-07-05T08:06:11.733495+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.11230v4","created_at":"2026-07-05T08:06:11.733495+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.11230","created_at":"2026-07-05T08:06:11.733495+00:00"},{"alias_kind":"pith_short_12","alias_value":"YTJKYBUESP7J","created_at":"2026-07-05T08:06:11.733495+00:00"},{"alias_kind":"pith_short_16","alias_value":"YTJKYBUESP7J6DKD","created_at":"2026-07-05T08:06:11.733495+00:00"},{"alias_kind":"pith_short_8","alias_value":"YTJKYBUE","created_at":"2026-07-05T08:06:11.733495+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.00387","citing_title":"From Objectives to Applications: Aligning Architectural Biases in Audio Self-Supervised Learning","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03169","citing_title":"SketchSong: Hierarchical Song Generation with Sketch Planning and Fine-Grained Multi-Track Modeling","ref_index":32,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB","json":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB.json","graph_json":"https://pith.science/api/pith-number/YTJKYBUESP7J6DKDYXGPXLEWWB/graph.json","events_json":"https://pith.science/api/pith-number/YTJKYBUESP7J6DKDYXGPXLEWWB/events.json","paper":"https://pith.science/paper/YTJKYBUE"},"agent_actions":{"view_html":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB","download_json":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB.json","view_paper":"https://pith.science/paper/YTJKYBUE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.11230&json=true","fetch_graph":"https://pith.science/api/pith-number/YTJKYBUESP7J6DKDYXGPXLEWWB/graph.json","fetch_events":"https://pith.science/api/pith-number/YTJKYBUESP7J6DKDYXGPXLEWWB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB/action/storage_attestation","attest_author":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB/action/author_attestation","sign_citation":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB/action/citation_signature","submit_replication":"https://pith.science/pith/YTJKYBUESP7J6DKDYXGPXLEWWB/action/replication_record"}},"created_at":"2026-07-05T08:06:11.733495+00:00","updated_at":"2026-07-05T08:06:11.733495+00:00"}