{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:SH5LB4K6WZT7PGJXGB5KCFSJ33","short_pith_number":"pith:SH5LB4K6","schema_version":"1.0","canonical_sha256":"91fab0f15eb667f79937307aa11649dee172b885f4a72b8b7c5670a50962b8f5","source":{"kind":"arxiv","id":"1908.08206","version":1},"attestation_state":"computed","paper":{"title":"Denoising based Sequence-to-Sequence Pre-training for Text Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jingming Liu, Liang Wang, Ruoyu Jia, Sujian Li, Wei Zhao","submitted_at":"2019-08-22T05:26:25Z","abstract_excerpt":"This paper presents a new sequence-to-sequence (seq2seq) pre-training method PoDA (Pre-training of Denoising Autoencoders), which learns representations suitable for text generation tasks. Unlike encoder-only (e.g., BERT) or decoder-only (e.g., OpenAI GPT) pre-training approaches, PoDA jointly pre-trains both the encoder and decoder by denoising the noise-corrupted text, and it also has the advantage of keeping the network architecture unchanged in the subsequent fine-tuning stage. Meanwhile, we design a hybrid model of Transformer and pointer-generator networks as the backbone architecture fo"},"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":"1908.08206","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2019-08-22T05:26:25Z","cross_cats_sorted":[],"title_canon_sha256":"bbb76c9ca334449f034aa1c67a99781031badbd349787bec54f6deb5256241ad","abstract_canon_sha256":"c3779cd8ee3f9443d6f4253d8e5e6d87dc6b11b051fb13bc1826f65a34f33d68"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-04T23:59:10.539153Z","signature_b64":"bWJgTJmKrp/nzulbGbHnCiXvLk4y3Llq5B86beGXX6NJUhPsHdDQWLIP08y2DbnbsX7wY2B6Z+LAaKa0O37vDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"91fab0f15eb667f79937307aa11649dee172b885f4a72b8b7c5670a50962b8f5","last_reissued_at":"2026-07-04T23:59:10.538810Z","signature_status":"signed_v1","first_computed_at":"2026-07-04T23:59:10.538810Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Denoising based Sequence-to-Sequence Pre-training for Text Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jingming Liu, Liang Wang, Ruoyu Jia, Sujian Li, Wei Zhao","submitted_at":"2019-08-22T05:26:25Z","abstract_excerpt":"This paper presents a new sequence-to-sequence (seq2seq) pre-training method PoDA (Pre-training of Denoising Autoencoders), which learns representations suitable for text generation tasks. Unlike encoder-only (e.g., BERT) or decoder-only (e.g., OpenAI GPT) pre-training approaches, PoDA jointly pre-trains both the encoder and decoder by denoising the noise-corrupted text, and it also has the advantage of keeping the network architecture unchanged in the subsequent fine-tuning stage. Meanwhile, we design a hybrid model of Transformer and pointer-generator networks as the backbone architecture fo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1908.08206","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/1908.08206/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":"1908.08206","created_at":"2026-07-04T23:59:10.538865+00:00"},{"alias_kind":"arxiv_version","alias_value":"1908.08206v1","created_at":"2026-07-04T23:59:10.538865+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1908.08206","created_at":"2026-07-04T23:59:10.538865+00:00"},{"alias_kind":"pith_short_12","alias_value":"SH5LB4K6WZT7","created_at":"2026-07-04T23:59:10.538865+00:00"},{"alias_kind":"pith_short_16","alias_value":"SH5LB4K6WZT7PGJX","created_at":"2026-07-04T23:59:10.538865+00:00"},{"alias_kind":"pith_short_8","alias_value":"SH5LB4K6","created_at":"2026-07-04T23:59:10.538865+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.08952","citing_title":"When LLM Agents Meet Graph Optimization: An Automated Data Quality Improvement Approach","ref_index":46,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33","json":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33.json","graph_json":"https://pith.science/api/pith-number/SH5LB4K6WZT7PGJXGB5KCFSJ33/graph.json","events_json":"https://pith.science/api/pith-number/SH5LB4K6WZT7PGJXGB5KCFSJ33/events.json","paper":"https://pith.science/paper/SH5LB4K6"},"agent_actions":{"view_html":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33","download_json":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33.json","view_paper":"https://pith.science/paper/SH5LB4K6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1908.08206&json=true","fetch_graph":"https://pith.science/api/pith-number/SH5LB4K6WZT7PGJXGB5KCFSJ33/graph.json","fetch_events":"https://pith.science/api/pith-number/SH5LB4K6WZT7PGJXGB5KCFSJ33/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33/action/storage_attestation","attest_author":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33/action/author_attestation","sign_citation":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33/action/citation_signature","submit_replication":"https://pith.science/pith/SH5LB4K6WZT7PGJXGB5KCFSJ33/action/replication_record"}},"created_at":"2026-07-04T23:59:10.538865+00:00","updated_at":"2026-07-04T23:59:10.538865+00:00"}