{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:OFSVSYOUROC7HPM337DTMMKPLG","short_pith_number":"pith:OFSVSYOU","schema_version":"1.0","canonical_sha256":"71655961d48b85f3bd9bdfc736314f59abbf8b44271b251db220cf9bf4854c11","source":{"kind":"arxiv","id":"2012.15525","version":3},"attestation_state":"computed","paper":{"title":"BANG: Bridging Autoregressive and Non-autoregressive Generation with Large Scale Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dayiheng Liu, Houqiang Li, Jian Jiao, Jiusheng Chen, Kewen Tang, Ming Zhou, Nan Duan, Ruofei Zhang, Weizhen Qi, Weizhu Chen, Yeyun Gong, Yu Yan","submitted_at":"2020-12-31T10:09:29Z","abstract_excerpt":"In this paper, we propose BANG, a new pretraining model to Bridge the gap between Autoregressive (AR) and Non-autoregressive (NAR) Generation. AR and NAR generation can be uniformly regarded as to what extent previous tokens can be attended, and BANG bridges AR and NAR generation by designing a novel model structure for large-scale pretraining. The pretrained BANG model can simultaneously support AR, NAR and semi-NAR generation to meet different requirements. Experiments on question generation (SQuAD 1.1), summarization (XSum) and dialogue generation (PersonaChat) show that BANG improves NAR a"},"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":"2012.15525","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-12-31T10:09:29Z","cross_cats_sorted":[],"title_canon_sha256":"b3afa496caeb4dd14fcaf79900fa682d488f40a0ab9a5e1a963ae3e55902d89e","abstract_canon_sha256":"2342689b4ed312bd78104700dc62dcefdb51267eb3f0347f4669e5a5f1d6c6db"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:49:52.601905Z","signature_b64":"tL+21lUA9VzdaGhFKGpmmkl0X05snd5an37vZj9lRNGeaSNUyUMxdWVHVWpfChqhgXDl93QpRqhz0FhZyNn5Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"71655961d48b85f3bd9bdfc736314f59abbf8b44271b251db220cf9bf4854c11","last_reissued_at":"2026-07-05T02:49:52.601421Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:49:52.601421Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BANG: Bridging Autoregressive and Non-autoregressive Generation with Large Scale Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dayiheng Liu, Houqiang Li, Jian Jiao, Jiusheng Chen, Kewen Tang, Ming Zhou, Nan Duan, Ruofei Zhang, Weizhen Qi, Weizhu Chen, Yeyun Gong, Yu Yan","submitted_at":"2020-12-31T10:09:29Z","abstract_excerpt":"In this paper, we propose BANG, a new pretraining model to Bridge the gap between Autoregressive (AR) and Non-autoregressive (NAR) Generation. AR and NAR generation can be uniformly regarded as to what extent previous tokens can be attended, and BANG bridges AR and NAR generation by designing a novel model structure for large-scale pretraining. The pretrained BANG model can simultaneously support AR, NAR and semi-NAR generation to meet different requirements. Experiments on question generation (SQuAD 1.1), summarization (XSum) and dialogue generation (PersonaChat) show that BANG improves NAR a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2012.15525","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/2012.15525/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":"2012.15525","created_at":"2026-07-05T02:49:52.601491+00:00"},{"alias_kind":"arxiv_version","alias_value":"2012.15525v3","created_at":"2026-07-05T02:49:52.601491+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2012.15525","created_at":"2026-07-05T02:49:52.601491+00:00"},{"alias_kind":"pith_short_12","alias_value":"OFSVSYOUROC7","created_at":"2026-07-05T02:49:52.601491+00:00"},{"alias_kind":"pith_short_16","alias_value":"OFSVSYOUROC7HPM3","created_at":"2026-07-05T02:49:52.601491+00:00"},{"alias_kind":"pith_short_8","alias_value":"OFSVSYOU","created_at":"2026-07-05T02:49:52.601491+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/OFSVSYOUROC7HPM337DTMMKPLG","json":"https://pith.science/pith/OFSVSYOUROC7HPM337DTMMKPLG.json","graph_json":"https://pith.science/api/pith-number/OFSVSYOUROC7HPM337DTMMKPLG/graph.json","events_json":"https://pith.science/api/pith-number/OFSVSYOUROC7HPM337DTMMKPLG/events.json","paper":"https://pith.science/paper/OFSVSYOU"},"agent_actions":{"view_html":"https://pith.science/pith/OFSVSYOUROC7HPM337DTMMKPLG","download_json":"https://pith.science/pith/OFSVSYOUROC7HPM337DTMMKPLG.json","view_paper":"https://pith.science/paper/OFSVSYOU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2012.15525&json=true","fetch_graph":"https://pith.science/api/pith-number/OFSVSYOUROC7HPM337DTMMKPLG/graph.json","fetch_events":"https://pith.science/api/pith-number/OFSVSYOUROC7HPM337DTMMKPLG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OFSVSYOUROC7HPM337DTMMKPLG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OFSVSYOUROC7HPM337DTMMKPLG/action/storage_attestation","attest_author":"https://pith.science/pith/OFSVSYOUROC7HPM337DTMMKPLG/action/author_attestation","sign_citation":"https://pith.science/pith/OFSVSYOUROC7HPM337DTMMKPLG/action/citation_signature","submit_replication":"https://pith.science/pith/OFSVSYOUROC7HPM337DTMMKPLG/action/replication_record"}},"created_at":"2026-07-05T02:49:52.601491+00:00","updated_at":"2026-07-05T02:49:52.601491+00:00"}