{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3OEX5GWLLVT4UCWXPLQOFGBPNX","short_pith_number":"pith:3OEX5GWL","schema_version":"1.0","canonical_sha256":"db897e9acb5d67ca0ad77ae0e2982f6dd471a1e574dc4b38b917460f2aa8c7a5","source":{"kind":"arxiv","id":"2411.17170","version":2},"attestation_state":"computed","paper":{"title":"Overcoming Non-monotonicity in Transducer-based Streaming Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Min Zhang, Yang Feng, Zhengrui Ma","submitted_at":"2024-11-26T07:19:26Z","abstract_excerpt":"Streaming generation models are utilized across fields, with the Transducer architecture being popular in industrial applications. However, its input-synchronous decoding mechanism presents challenges in tasks requiring non-monotonic alignments, such as simultaneous translation. In this research, we address this issue by integrating Transducer's decoding with the history of input stream via a learnable monotonic attention. Our approach leverages the forward-backward algorithm to infer the posterior probability of alignments between the predictor states and input timestamps, which is then used "},"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":"2411.17170","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-11-26T07:19:26Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"529ddc0137d5f19de6bb828f1c6df9c542a712ca51a180dd963a96f02b14b8ad","abstract_canon_sha256":"07b176de8d71c68578f372566d258d441de067658f53cb98b7006f1fe839910e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:10:49.123456Z","signature_b64":"7kHVB3qbW6Jx3rRGW67OU5K3BsVTBXSjbrCBBYNVFonuXWJXay3aTaOE7fWnHLoWRxzk2cok4v9WFlMjnRRTAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db897e9acb5d67ca0ad77ae0e2982f6dd471a1e574dc4b38b917460f2aa8c7a5","last_reissued_at":"2026-07-05T11:10:49.122948Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:10:49.122948Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Overcoming Non-monotonicity in Transducer-based Streaming Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Min Zhang, Yang Feng, Zhengrui Ma","submitted_at":"2024-11-26T07:19:26Z","abstract_excerpt":"Streaming generation models are utilized across fields, with the Transducer architecture being popular in industrial applications. However, its input-synchronous decoding mechanism presents challenges in tasks requiring non-monotonic alignments, such as simultaneous translation. In this research, we address this issue by integrating Transducer's decoding with the history of input stream via a learnable monotonic attention. Our approach leverages the forward-backward algorithm to infer the posterior probability of alignments between the predictor states and input timestamps, which is then used "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.17170","kind":"arxiv","version":2},"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/2411.17170/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":"2411.17170","created_at":"2026-07-05T11:10:49.123004+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.17170v2","created_at":"2026-07-05T11:10:49.123004+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.17170","created_at":"2026-07-05T11:10:49.123004+00:00"},{"alias_kind":"pith_short_12","alias_value":"3OEX5GWLLVT4","created_at":"2026-07-05T11:10:49.123004+00:00"},{"alias_kind":"pith_short_16","alias_value":"3OEX5GWLLVT4UCWX","created_at":"2026-07-05T11:10:49.123004+00:00"},{"alias_kind":"pith_short_8","alias_value":"3OEX5GWL","created_at":"2026-07-05T11:10:49.123004+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.07803","citing_title":"StreamUni: Achieving Streaming Speech Translation with a Unified Large Speech-Language Model","ref_index":25,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX","json":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX.json","graph_json":"https://pith.science/api/pith-number/3OEX5GWLLVT4UCWXPLQOFGBPNX/graph.json","events_json":"https://pith.science/api/pith-number/3OEX5GWLLVT4UCWXPLQOFGBPNX/events.json","paper":"https://pith.science/paper/3OEX5GWL"},"agent_actions":{"view_html":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX","download_json":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX.json","view_paper":"https://pith.science/paper/3OEX5GWL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.17170&json=true","fetch_graph":"https://pith.science/api/pith-number/3OEX5GWLLVT4UCWXPLQOFGBPNX/graph.json","fetch_events":"https://pith.science/api/pith-number/3OEX5GWLLVT4UCWXPLQOFGBPNX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX/action/storage_attestation","attest_author":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX/action/author_attestation","sign_citation":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX/action/citation_signature","submit_replication":"https://pith.science/pith/3OEX5GWLLVT4UCWXPLQOFGBPNX/action/replication_record"}},"created_at":"2026-07-05T11:10:49.123004+00:00","updated_at":"2026-07-05T11:10:49.123004+00:00"}