{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:FN7EBROXEOJALQYQIPPZMSWXQC","short_pith_number":"pith:FN7EBROX","schema_version":"1.0","canonical_sha256":"2b7e40c5d7239205c31043df964ad7809bc17c717bed1817e98aaea76fde6d08","source":{"kind":"arxiv","id":"2004.07437","version":3},"attestation_state":"computed","paper":{"title":"Non-Autoregressive Machine Translation with Latent Alignments","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chitwan Saharia, Mohammad Norouzi, Saurabh Saxena, William Chan","submitted_at":"2020-04-16T03:45:56Z","abstract_excerpt":"This paper presents two strong methods, CTC and Imputer, for non-autoregressive machine translation that model latent alignments with dynamic programming. We revisit CTC for machine translation and demonstrate that a simple CTC model can achieve state-of-the-art for single-step non-autoregressive machine translation, contrary to what prior work indicates. In addition, we adapt the Imputer model for non-autoregressive machine translation and demonstrate that Imputer with just 4 generation steps can match the performance of an autoregressive Transformer baseline. Our latent alignment models are "},"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":"2004.07437","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-16T03:45:56Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"12552d480db5880caf42b85038e8e17ff59fd99b919cb226127e9df0d090a5a9","abstract_canon_sha256":"9c35b38624dc642f45949a67a3658592530edb9ed1b077c1de92e4c48fbc7cce"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:51:40.762701Z","signature_b64":"OnPMvhAvfCBse9Qv24rqmen4ztOu5pOWaPzZpCL/3LwSGIc06Y7BbTlQJwNIk3JTp12I7zT2LYbCoyRak0vBDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2b7e40c5d7239205c31043df964ad7809bc17c717bed1817e98aaea76fde6d08","last_reissued_at":"2026-07-05T01:51:40.762258Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:51:40.762258Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Non-Autoregressive Machine Translation with Latent Alignments","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Chitwan Saharia, Mohammad Norouzi, Saurabh Saxena, William Chan","submitted_at":"2020-04-16T03:45:56Z","abstract_excerpt":"This paper presents two strong methods, CTC and Imputer, for non-autoregressive machine translation that model latent alignments with dynamic programming. We revisit CTC for machine translation and demonstrate that a simple CTC model can achieve state-of-the-art for single-step non-autoregressive machine translation, contrary to what prior work indicates. In addition, we adapt the Imputer model for non-autoregressive machine translation and demonstrate that Imputer with just 4 generation steps can match the performance of an autoregressive Transformer baseline. Our latent alignment models are "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.07437","kind":"arxiv","version":3},"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/2004.07437/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":"2004.07437","created_at":"2026-07-05T01:51:40.762325+00:00"},{"alias_kind":"arxiv_version","alias_value":"2004.07437v3","created_at":"2026-07-05T01:51:40.762325+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.07437","created_at":"2026-07-05T01:51:40.762325+00:00"},{"alias_kind":"pith_short_12","alias_value":"FN7EBROXEOJA","created_at":"2026-07-05T01:51:40.762325+00:00"},{"alias_kind":"pith_short_16","alias_value":"FN7EBROXEOJALQYQ","created_at":"2026-07-05T01:51:40.762325+00:00"},{"alias_kind":"pith_short_8","alias_value":"FN7EBROX","created_at":"2026-07-05T01:51:40.762325+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2211.15089","citing_title":"Continuous diffusion for categorical data","ref_index":74,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC","json":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC.json","graph_json":"https://pith.science/api/pith-number/FN7EBROXEOJALQYQIPPZMSWXQC/graph.json","events_json":"https://pith.science/api/pith-number/FN7EBROXEOJALQYQIPPZMSWXQC/events.json","paper":"https://pith.science/paper/FN7EBROX"},"agent_actions":{"view_html":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC","download_json":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC.json","view_paper":"https://pith.science/paper/FN7EBROX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2004.07437&json=true","fetch_graph":"https://pith.science/api/pith-number/FN7EBROXEOJALQYQIPPZMSWXQC/graph.json","fetch_events":"https://pith.science/api/pith-number/FN7EBROXEOJALQYQIPPZMSWXQC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC/action/storage_attestation","attest_author":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC/action/author_attestation","sign_citation":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC/action/citation_signature","submit_replication":"https://pith.science/pith/FN7EBROXEOJALQYQIPPZMSWXQC/action/replication_record"}},"created_at":"2026-07-05T01:51:40.762325+00:00","updated_at":"2026-07-05T01:51:40.762325+00:00"}