{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:L2ZN2TV5H24B35UCNPG7J5ZCC7","short_pith_number":"pith:L2ZN2TV5","schema_version":"1.0","canonical_sha256":"5eb2dd4ebd3eb81df6826bcdf4f72217cbdd072a167a8ea61dc00b6c6ae7eca0","source":{"kind":"arxiv","id":"1911.03829","version":3},"attestation_state":"computed","paper":{"title":"Distilling Knowledge Learned in BERT for Text Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jingjing Liu, Jingzhou Liu, Yen-Chun Chen, Yu Cheng, Zhe Gan","submitted_at":"2019-11-10T02:12:38Z","abstract_excerpt":"Large-scale pre-trained language model such as BERT has achieved great success in language understanding tasks. However, it remains an open question how to utilize BERT for language generation. In this paper, we present a novel approach, Conditional Masked Language Modeling (C-MLM), to enable the finetuning of BERT on target generation tasks. The finetuned BERT (teacher) is exploited as extra supervision to improve conventional Seq2Seq models (student) for better text generation performance. By leveraging BERT's idiosyncratic bidirectional nature, distilling knowledge learned in BERT can encou"},"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":"1911.03829","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-11-10T02:12:38Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"1a5940533b2adbd70279f0afa1ecba367772cf68e4d382839dc803cce5a143b8","abstract_canon_sha256":"2cd3cd4729ad22c7286d0af917c19f5884edcf11de50207010b5b49de4786385"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:20:13.357761Z","signature_b64":"lr5rDAUzfslzbYBqs4+mGfDca9ZA5RK1fkzuZfkBybTKZ5Nw1x4SqSGAHE5p2qpB9MqA/XlZSqHenAalmPwBCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5eb2dd4ebd3eb81df6826bcdf4f72217cbdd072a167a8ea61dc00b6c6ae7eca0","last_reissued_at":"2026-07-05T01:20:13.357249Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:20:13.357249Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Distilling Knowledge Learned in BERT for Text Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jingjing Liu, Jingzhou Liu, Yen-Chun Chen, Yu Cheng, Zhe Gan","submitted_at":"2019-11-10T02:12:38Z","abstract_excerpt":"Large-scale pre-trained language model such as BERT has achieved great success in language understanding tasks. However, it remains an open question how to utilize BERT for language generation. In this paper, we present a novel approach, Conditional Masked Language Modeling (C-MLM), to enable the finetuning of BERT on target generation tasks. The finetuned BERT (teacher) is exploited as extra supervision to improve conventional Seq2Seq models (student) for better text generation performance. By leveraging BERT's idiosyncratic bidirectional nature, distilling knowledge learned in BERT can encou"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1911.03829","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/1911.03829/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":"1911.03829","created_at":"2026-07-05T01:20:13.357308+00:00"},{"alias_kind":"arxiv_version","alias_value":"1911.03829v3","created_at":"2026-07-05T01:20:13.357308+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1911.03829","created_at":"2026-07-05T01:20:13.357308+00:00"},{"alias_kind":"pith_short_12","alias_value":"L2ZN2TV5H24B","created_at":"2026-07-05T01:20:13.357308+00:00"},{"alias_kind":"pith_short_16","alias_value":"L2ZN2TV5H24B35UC","created_at":"2026-07-05T01:20:13.357308+00:00"},{"alias_kind":"pith_short_8","alias_value":"L2ZN2TV5","created_at":"2026-07-05T01:20:13.357308+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/L2ZN2TV5H24B35UCNPG7J5ZCC7","json":"https://pith.science/pith/L2ZN2TV5H24B35UCNPG7J5ZCC7.json","graph_json":"https://pith.science/api/pith-number/L2ZN2TV5H24B35UCNPG7J5ZCC7/graph.json","events_json":"https://pith.science/api/pith-number/L2ZN2TV5H24B35UCNPG7J5ZCC7/events.json","paper":"https://pith.science/paper/L2ZN2TV5"},"agent_actions":{"view_html":"https://pith.science/pith/L2ZN2TV5H24B35UCNPG7J5ZCC7","download_json":"https://pith.science/pith/L2ZN2TV5H24B35UCNPG7J5ZCC7.json","view_paper":"https://pith.science/paper/L2ZN2TV5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1911.03829&json=true","fetch_graph":"https://pith.science/api/pith-number/L2ZN2TV5H24B35UCNPG7J5ZCC7/graph.json","fetch_events":"https://pith.science/api/pith-number/L2ZN2TV5H24B35UCNPG7J5ZCC7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L2ZN2TV5H24B35UCNPG7J5ZCC7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L2ZN2TV5H24B35UCNPG7J5ZCC7/action/storage_attestation","attest_author":"https://pith.science/pith/L2ZN2TV5H24B35UCNPG7J5ZCC7/action/author_attestation","sign_citation":"https://pith.science/pith/L2ZN2TV5H24B35UCNPG7J5ZCC7/action/citation_signature","submit_replication":"https://pith.science/pith/L2ZN2TV5H24B35UCNPG7J5ZCC7/action/replication_record"}},"created_at":"2026-07-05T01:20:13.357308+00:00","updated_at":"2026-07-05T01:20:13.357308+00:00"}