{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:X544VY3TJLHX3B4UMU7LB4W6E3","short_pith_number":"pith:X544VY3T","schema_version":"1.0","canonical_sha256":"bf79cae3734acf7d8794653eb0f2de26cc1a2a3f3b5b1b015b1647584745c49a","source":{"kind":"arxiv","id":"2205.14333","version":1},"attestation_state":"computed","paper":{"title":"One Reference Is Not Enough: Diverse Distillation with Reference Selection for Non-Autoregressive Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chenze Shao, Xuanfu Wu, Yang Feng","submitted_at":"2022-05-28T04:59:33Z","abstract_excerpt":"Non-autoregressive neural machine translation (NAT) suffers from the multi-modality problem: the source sentence may have multiple correct translations, but the loss function is calculated only according to the reference sentence. Sequence-level knowledge distillation makes the target more deterministic by replacing the target with the output from an autoregressive model. However, the multi-modality problem in the distilled dataset is still nonnegligible. Furthermore, learning from a specific teacher limits the upper bound of the model capability, restricting the potential of NAT models. In th"},"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":"2205.14333","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-28T04:59:33Z","cross_cats_sorted":[],"title_canon_sha256":"fbd06b8ef5bedeaf59ee025989dcd90a6ca92be459ecf08f6e56dc0880dfa71c","abstract_canon_sha256":"deaabba0551b879c1e29859fbcb6c9f8b7e8fb91ca5e8f82e49056005e6c010e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:27:16.332917Z","signature_b64":"FGEAEBDjh1xJauFLXy9KS7LZcBWld1ObKAiEG/tnBCu60AckcrscvJVXeMSkUCSQs5/znyzr2SF6fqGiFfrDDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bf79cae3734acf7d8794653eb0f2de26cc1a2a3f3b5b1b015b1647584745c49a","last_reissued_at":"2026-07-05T04:27:16.332483Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:27:16.332483Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"One Reference Is Not Enough: Diverse Distillation with Reference Selection for Non-Autoregressive Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chenze Shao, Xuanfu Wu, Yang Feng","submitted_at":"2022-05-28T04:59:33Z","abstract_excerpt":"Non-autoregressive neural machine translation (NAT) suffers from the multi-modality problem: the source sentence may have multiple correct translations, but the loss function is calculated only according to the reference sentence. Sequence-level knowledge distillation makes the target more deterministic by replacing the target with the output from an autoregressive model. However, the multi-modality problem in the distilled dataset is still nonnegligible. Furthermore, learning from a specific teacher limits the upper bound of the model capability, restricting the potential of NAT models. In th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.14333","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/2205.14333/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":"2205.14333","created_at":"2026-07-05T04:27:16.332542+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.14333v1","created_at":"2026-07-05T04:27:16.332542+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.14333","created_at":"2026-07-05T04:27:16.332542+00:00"},{"alias_kind":"pith_short_12","alias_value":"X544VY3TJLHX","created_at":"2026-07-05T04:27:16.332542+00:00"},{"alias_kind":"pith_short_16","alias_value":"X544VY3TJLHX3B4U","created_at":"2026-07-05T04:27:16.332542+00:00"},{"alias_kind":"pith_short_8","alias_value":"X544VY3T","created_at":"2026-07-05T04:27:16.332542+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.13485","citing_title":"Curriculum Learning for Biological Sequence Prediction: The Case of De Novo Peptide Sequencing","ref_index":42,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3","json":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3.json","graph_json":"https://pith.science/api/pith-number/X544VY3TJLHX3B4UMU7LB4W6E3/graph.json","events_json":"https://pith.science/api/pith-number/X544VY3TJLHX3B4UMU7LB4W6E3/events.json","paper":"https://pith.science/paper/X544VY3T"},"agent_actions":{"view_html":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3","download_json":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3.json","view_paper":"https://pith.science/paper/X544VY3T","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.14333&json=true","fetch_graph":"https://pith.science/api/pith-number/X544VY3TJLHX3B4UMU7LB4W6E3/graph.json","fetch_events":"https://pith.science/api/pith-number/X544VY3TJLHX3B4UMU7LB4W6E3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3/action/storage_attestation","attest_author":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3/action/author_attestation","sign_citation":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3/action/citation_signature","submit_replication":"https://pith.science/pith/X544VY3TJLHX3B4UMU7LB4W6E3/action/replication_record"}},"created_at":"2026-07-05T04:27:16.332542+00:00","updated_at":"2026-07-05T04:27:16.332542+00:00"}