{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:SLUFLAAYXZWOT777YHSTLQNMSV","short_pith_number":"pith:SLUFLAAY","schema_version":"1.0","canonical_sha256":"92e8558018be6ce9ffffc1e535c1ac9552f0225a4aca4a3c40e68b8a1e8d4118","source":{"kind":"arxiv","id":"2210.03953","version":1},"attestation_state":"computed","paper":{"title":"Non-Monotonic Latent Alignments for CTC-Based Non-Autoregressive Machine Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chenze Shao, Yang Feng","submitted_at":"2022-10-08T07:44:28Z","abstract_excerpt":"Non-autoregressive translation (NAT) models are typically trained with the cross-entropy loss, which forces the model outputs to be aligned verbatim with the target sentence and will highly penalize small shifts in word positions. Latent alignment models relax the explicit alignment by marginalizing out all monotonic latent alignments with the CTC loss. However, they cannot handle non-monotonic alignments, which is non-negligible as there is typically global word reordering in machine translation. In this work, we explore non-monotonic latent alignments for NAT. We extend the alignment space t"},"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":"2210.03953","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-10-08T07:44:28Z","cross_cats_sorted":[],"title_canon_sha256":"e0a9952a110bcf2a806e511673d14370e74eb3c1a66db1710146016bed9aa04a","abstract_canon_sha256":"137442882096acdbaa0d69b8b1290b3b72185d6c87612b7371b0c80bdf80c6ad"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:04:34.306120Z","signature_b64":"sXFes+hQbrGfPsUFqrY5ACU21pHSNWQczAkpEgxEnmIV7U9DLnExfIiHhxFg2PmLhlUIC/uwIzX73i5yhhI1Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"92e8558018be6ce9ffffc1e535c1ac9552f0225a4aca4a3c40e68b8a1e8d4118","last_reissued_at":"2026-07-05T05:04:34.305698Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:04:34.305698Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Non-Monotonic Latent Alignments for CTC-Based Non-Autoregressive Machine Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chenze Shao, Yang Feng","submitted_at":"2022-10-08T07:44:28Z","abstract_excerpt":"Non-autoregressive translation (NAT) models are typically trained with the cross-entropy loss, which forces the model outputs to be aligned verbatim with the target sentence and will highly penalize small shifts in word positions. Latent alignment models relax the explicit alignment by marginalizing out all monotonic latent alignments with the CTC loss. However, they cannot handle non-monotonic alignments, which is non-negligible as there is typically global word reordering in machine translation. In this work, we explore non-monotonic latent alignments for NAT. We extend the alignment space t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.03953","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/2210.03953/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":"2210.03953","created_at":"2026-07-05T05:04:34.305756+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.03953v1","created_at":"2026-07-05T05:04:34.305756+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.03953","created_at":"2026-07-05T05:04:34.305756+00:00"},{"alias_kind":"pith_short_12","alias_value":"SLUFLAAYXZWO","created_at":"2026-07-05T05:04:34.305756+00:00"},{"alias_kind":"pith_short_16","alias_value":"SLUFLAAYXZWOT777","created_at":"2026-07-05T05:04:34.305756+00:00"},{"alias_kind":"pith_short_8","alias_value":"SLUFLAAY","created_at":"2026-07-05T05:04:34.305756+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.17170","citing_title":"Overcoming Non-monotonicity in Transducer-based Streaming Generation","ref_index":38,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV","json":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV.json","graph_json":"https://pith.science/api/pith-number/SLUFLAAYXZWOT777YHSTLQNMSV/graph.json","events_json":"https://pith.science/api/pith-number/SLUFLAAYXZWOT777YHSTLQNMSV/events.json","paper":"https://pith.science/paper/SLUFLAAY"},"agent_actions":{"view_html":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV","download_json":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV.json","view_paper":"https://pith.science/paper/SLUFLAAY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.03953&json=true","fetch_graph":"https://pith.science/api/pith-number/SLUFLAAYXZWOT777YHSTLQNMSV/graph.json","fetch_events":"https://pith.science/api/pith-number/SLUFLAAYXZWOT777YHSTLQNMSV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV/action/storage_attestation","attest_author":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV/action/author_attestation","sign_citation":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV/action/citation_signature","submit_replication":"https://pith.science/pith/SLUFLAAYXZWOT777YHSTLQNMSV/action/replication_record"}},"created_at":"2026-07-05T05:04:34.305756+00:00","updated_at":"2026-07-05T05:04:34.305756+00:00"}