{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:3E3NB2DWKA5JDFZIAV35DYHK2S","short_pith_number":"pith:3E3NB2DW","schema_version":"1.0","canonical_sha256":"d936d0e876503a9197280577d1e0ead4a898ef0ca98601bcf7b4e012e40ffaf4","source":{"kind":"arxiv","id":"2107.13689","version":1},"attestation_state":"computed","paper":{"title":"Using Perturbed Length-aware Positional Encoding for Non-autoregressive Neural Machine Translation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Katsuhito Sudoh, Satoshi Nakamura, Yui Oka","submitted_at":"2021-07-29T00:51:44Z","abstract_excerpt":"Non-autoregressive neural machine translation (NAT) usually employs sequence-level knowledge distillation using autoregressive neural machine translation (AT) as its teacher model. However, a NAT model often outputs shorter sentences than an AT model. In this work, we propose sequence-level knowledge distillation (SKD) using perturbed length-aware positional encoding and apply it to a student model, the Levenshtein Transformer. Our method outperformed a standard Levenshtein Transformer by 2.5 points in bilingual evaluation understudy (BLEU) at maximum in a WMT14 German to English translation. "},"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":"2107.13689","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2021-07-29T00:51:44Z","cross_cats_sorted":[],"title_canon_sha256":"0bb5b7420ae83e0ef8494dc64629aeb757f03acac9e06e51625f6c97cd5c21be","abstract_canon_sha256":"611310a9e6bf48041e318ca4ebb1108166ef846829933744b4e12c166fa05ff9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:01:39.212369Z","signature_b64":"UhLM3PKlX2f5LyxdERIj6jDlsHltOU772iN6/g3Pdfmh1rMwCDz4VtO+x2N3ubZsu3Pt9k/UtThzgMvEtiVJDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d936d0e876503a9197280577d1e0ead4a898ef0ca98601bcf7b4e012e40ffaf4","last_reissued_at":"2026-07-05T03:01:39.212020Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:01:39.212020Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Using Perturbed Length-aware Positional Encoding for Non-autoregressive Neural Machine Translation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Katsuhito Sudoh, Satoshi Nakamura, Yui Oka","submitted_at":"2021-07-29T00:51:44Z","abstract_excerpt":"Non-autoregressive neural machine translation (NAT) usually employs sequence-level knowledge distillation using autoregressive neural machine translation (AT) as its teacher model. However, a NAT model often outputs shorter sentences than an AT model. In this work, we propose sequence-level knowledge distillation (SKD) using perturbed length-aware positional encoding and apply it to a student model, the Levenshtein Transformer. Our method outperformed a standard Levenshtein Transformer by 2.5 points in bilingual evaluation understudy (BLEU) at maximum in a WMT14 German to English translation. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2107.13689","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/2107.13689/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":"2107.13689","created_at":"2026-07-05T03:01:39.212075+00:00"},{"alias_kind":"arxiv_version","alias_value":"2107.13689v1","created_at":"2026-07-05T03:01:39.212075+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2107.13689","created_at":"2026-07-05T03:01:39.212075+00:00"},{"alias_kind":"pith_short_12","alias_value":"3E3NB2DWKA5J","created_at":"2026-07-05T03:01:39.212075+00:00"},{"alias_kind":"pith_short_16","alias_value":"3E3NB2DWKA5JDFZI","created_at":"2026-07-05T03:01:39.212075+00:00"},{"alias_kind":"pith_short_8","alias_value":"3E3NB2DW","created_at":"2026-07-05T03:01:39.212075+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S","json":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S.json","graph_json":"https://pith.science/api/pith-number/3E3NB2DWKA5JDFZIAV35DYHK2S/graph.json","events_json":"https://pith.science/api/pith-number/3E3NB2DWKA5JDFZIAV35DYHK2S/events.json","paper":"https://pith.science/paper/3E3NB2DW"},"agent_actions":{"view_html":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S","download_json":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S.json","view_paper":"https://pith.science/paper/3E3NB2DW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2107.13689&json=true","fetch_graph":"https://pith.science/api/pith-number/3E3NB2DWKA5JDFZIAV35DYHK2S/graph.json","fetch_events":"https://pith.science/api/pith-number/3E3NB2DWKA5JDFZIAV35DYHK2S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S/action/storage_attestation","attest_author":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S/action/author_attestation","sign_citation":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S/action/citation_signature","submit_replication":"https://pith.science/pith/3E3NB2DWKA5JDFZIAV35DYHK2S/action/replication_record"}},"created_at":"2026-07-05T03:01:39.212075+00:00","updated_at":"2026-07-05T03:01:39.212075+00:00"}