{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:R3MSXPALHVLHKAXUHSZXC2HYOJ","short_pith_number":"pith:R3MSXPAL","schema_version":"1.0","canonical_sha256":"8ed92bbc0b3d567502f43cb37168f8726d86a0c804c3b8e6a769a71c4115ba7a","source":{"kind":"arxiv","id":"1808.07561","version":2},"attestation_state":"computed","paper":{"title":"Training Deeper Neural Machine Translation Models with Transparent Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Ankur Bapna, Mia Xu Chen, Orhan Firat, Yonghui Wu, Yuan Cao","submitted_at":"2018-08-22T20:53:37Z","abstract_excerpt":"While current state-of-the-art NMT models, such as RNN seq2seq and Transformers, possess a large number of parameters, they are still shallow in comparison to convolutional models used for both text and vision applications. In this work we attempt to train significantly (2-3x) deeper Transformer and Bi-RNN encoders for machine translation. We propose a simple modification to the attention mechanism that eases the optimization of deeper models, and results in consistent gains of 0.7-1.1 BLEU on the benchmark WMT'14 English-German and WMT'15 Czech-English tasks for both architectures."},"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":"1808.07561","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-22T20:53:37Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"660dee33129ed3479f463915e54e55d33dca5516abfe1c53fe440c9a2741c2d1","abstract_canon_sha256":"39a5588aa82c663e28517d9e0c9ede1fa18275254acd1d418c94f81bc014aa64"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:06:31.506356Z","signature_b64":"7i/ft3JBQLwFQvgAdRqQzgPBcDo8xY+n7HW/80Uegt2+uKZmqhxY8GVe0NBGdW4eFBOla9xjAoY0379XQmGZBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8ed92bbc0b3d567502f43cb37168f8726d86a0c804c3b8e6a769a71c4115ba7a","last_reissued_at":"2026-05-18T00:06:31.505935Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:06:31.505935Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training Deeper Neural Machine Translation Models with Transparent Attention","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Ankur Bapna, Mia Xu Chen, Orhan Firat, Yonghui Wu, Yuan Cao","submitted_at":"2018-08-22T20:53:37Z","abstract_excerpt":"While current state-of-the-art NMT models, such as RNN seq2seq and Transformers, possess a large number of parameters, they are still shallow in comparison to convolutional models used for both text and vision applications. In this work we attempt to train significantly (2-3x) deeper Transformer and Bi-RNN encoders for machine translation. We propose a simple modification to the attention mechanism that eases the optimization of deeper models, and results in consistent gains of 0.7-1.1 BLEU on the benchmark WMT'14 English-German and WMT'15 Czech-English tasks for both architectures."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.07561","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"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":"1808.07561","created_at":"2026-05-18T00:06:31.506009+00:00"},{"alias_kind":"arxiv_version","alias_value":"1808.07561v2","created_at":"2026-05-18T00:06:31.506009+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.07561","created_at":"2026-05-18T00:06:31.506009+00:00"},{"alias_kind":"pith_short_12","alias_value":"R3MSXPALHVLH","created_at":"2026-05-18T12:32:50.500415+00:00"},{"alias_kind":"pith_short_16","alias_value":"R3MSXPALHVLHKAXU","created_at":"2026-05-18T12:32:50.500415+00:00"},{"alias_kind":"pith_short_8","alias_value":"R3MSXPAL","created_at":"2026-05-18T12:32:50.500415+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"1906.10770","citing_title":"Deep Modular Co-Attention Networks for Visual Question Answering","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"1906.12284","citing_title":"Widening the Representation Bottleneck in Neural Machine Translation with Lexical Shortcuts","ref_index":1,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ","json":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ.json","graph_json":"https://pith.science/api/pith-number/R3MSXPALHVLHKAXUHSZXC2HYOJ/graph.json","events_json":"https://pith.science/api/pith-number/R3MSXPALHVLHKAXUHSZXC2HYOJ/events.json","paper":"https://pith.science/paper/R3MSXPAL"},"agent_actions":{"view_html":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ","download_json":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ.json","view_paper":"https://pith.science/paper/R3MSXPAL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1808.07561&json=true","fetch_graph":"https://pith.science/api/pith-number/R3MSXPALHVLHKAXUHSZXC2HYOJ/graph.json","fetch_events":"https://pith.science/api/pith-number/R3MSXPALHVLHKAXUHSZXC2HYOJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ/action/storage_attestation","attest_author":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ/action/author_attestation","sign_citation":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ/action/citation_signature","submit_replication":"https://pith.science/pith/R3MSXPALHVLHKAXUHSZXC2HYOJ/action/replication_record"}},"created_at":"2026-05-18T00:06:31.506009+00:00","updated_at":"2026-05-18T00:06:31.506009+00:00"}