{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:GCIOQXNNZBIWUY2B2P34JPWDCD","short_pith_number":"pith:GCIOQXNN","canonical_record":{"source":{"id":"2009.14658","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-09-30T13:18:57Z","cross_cats_sorted":[],"title_canon_sha256":"b62b720952f6ac31d0deacae62bc027f1ec48eeddd695acd2119524827ef8f36","abstract_canon_sha256":"a682c6b6887eb4619e1404cea25876087d5b2fe57b2311b198d4f050e464fd08"},"schema_version":"1.0"},"canonical_sha256":"3090e85dadc8516a6341d3f7c4bec310e20d0db6d438046ee81a40cc037c90c1","source":{"kind":"arxiv","id":"2009.14658","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2009.14658","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"arxiv_version","alias_value":"2009.14658v2","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2009.14658","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"pith_short_12","alias_value":"GCIOQXNNZBIW","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"pith_short_16","alias_value":"GCIOQXNNZBIWUY2B","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"pith_short_8","alias_value":"GCIOQXNN","created_at":"2026-07-05T03:13:00Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:GCIOQXNNZBIWUY2B2P34JPWDCD","target":"record","payload":{"canonical_record":{"source":{"id":"2009.14658","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-09-30T13:18:57Z","cross_cats_sorted":[],"title_canon_sha256":"b62b720952f6ac31d0deacae62bc027f1ec48eeddd695acd2119524827ef8f36","abstract_canon_sha256":"a682c6b6887eb4619e1404cea25876087d5b2fe57b2311b198d4f050e464fd08"},"schema_version":"1.0"},"canonical_sha256":"3090e85dadc8516a6341d3f7c4bec310e20d0db6d438046ee81a40cc037c90c1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:13:00.458079Z","signature_b64":"KISPlFChe7m/wXDVyiwWjfzrB/PKQWQIfNwvsFy+hpRBiw+VQ1mMrNRW9OLODXFcf4rQ8UEjiEmETMHa06SaCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3090e85dadc8516a6341d3f7c4bec310e20d0db6d438046ee81a40cc037c90c1","last_reissued_at":"2026-07-05T03:13:00.457723Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:13:00.457723Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2009.14658","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T03:13:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E81iMOXf+lF6tQc1Pq0vA+2SSoCs8rPZ/t5b8o2XI49UlHxgRk16lS4EoFgnQcJONB47m8bHNCo7GqWZBZz3DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-08T13:57:54.614627Z"},"content_sha256":"d21093340a65c13a5a911505108b99525a59992b1fd3c493bb59f27b35dc17ef","schema_version":"1.0","event_id":"sha256:d21093340a65c13a5a911505108b99525a59992b1fd3c493bb59f27b35dc17ef"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:GCIOQXNNZBIWUY2B2P34JPWDCD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Learning Hard Retrieval Decoder Attention for Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Deyi Xiong, Hongfei Xu, Josef van Genabith, Qiuhui Liu","submitted_at":"2020-09-30T13:18:57Z","abstract_excerpt":"The Transformer translation model is based on the multi-head attention mechanism, which can be parallelized easily. The multi-head attention network performs the scaled dot-product attention function in parallel, empowering the model by jointly attending to information from different representation subspaces at different positions. In this paper, we present an approach to learning a hard retrieval attention where an attention head only attends to one token in the sentence rather than all tokens. The matrix multiplication between attention probabilities and the value sequence in the standard sc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2009.14658","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2009.14658/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T03:13:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sG2papgBawm/kKsXN1GvqNyWxp83zDQmRe2fjIrT4U86QJtl3AgKLv551pYyP+EHcCH6COhVufQAlyxQLLtbCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-08T13:57:54.615273Z"},"content_sha256":"457d953e0cdc16373200678af06c0704441028ca975ef53fab8c3dd211b18822","schema_version":"1.0","event_id":"sha256:457d953e0cdc16373200678af06c0704441028ca975ef53fab8c3dd211b18822"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GCIOQXNNZBIWUY2B2P34JPWDCD/bundle.json","state_url":"https://pith.science/pith/GCIOQXNNZBIWUY2B2P34JPWDCD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GCIOQXNNZBIWUY2B2P34JPWDCD/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-07-08T13:57:54Z","links":{"resolver":"https://pith.science/pith/GCIOQXNNZBIWUY2B2P34JPWDCD","bundle":"https://pith.science/pith/GCIOQXNNZBIWUY2B2P34JPWDCD/bundle.json","state":"https://pith.science/pith/GCIOQXNNZBIWUY2B2P34JPWDCD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GCIOQXNNZBIWUY2B2P34JPWDCD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:GCIOQXNNZBIWUY2B2P34JPWDCD","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"a682c6b6887eb4619e1404cea25876087d5b2fe57b2311b198d4f050e464fd08","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-09-30T13:18:57Z","title_canon_sha256":"b62b720952f6ac31d0deacae62bc027f1ec48eeddd695acd2119524827ef8f36"},"schema_version":"1.0","source":{"id":"2009.14658","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2009.14658","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"arxiv_version","alias_value":"2009.14658v2","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2009.14658","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"pith_short_12","alias_value":"GCIOQXNNZBIW","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"pith_short_16","alias_value":"GCIOQXNNZBIWUY2B","created_at":"2026-07-05T03:13:00Z"},{"alias_kind":"pith_short_8","alias_value":"GCIOQXNN","created_at":"2026-07-05T03:13:00Z"}],"graph_snapshots":[{"event_id":"sha256:457d953e0cdc16373200678af06c0704441028ca975ef53fab8c3dd211b18822","target":"graph","created_at":"2026-07-05T03:13:00Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2009.14658/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The Transformer translation model is based on the multi-head attention mechanism, which can be parallelized easily. The multi-head attention network performs the scaled dot-product attention function in parallel, empowering the model by jointly attending to information from different representation subspaces at different positions. In this paper, we present an approach to learning a hard retrieval attention where an attention head only attends to one token in the sentence rather than all tokens. The matrix multiplication between attention probabilities and the value sequence in the standard sc","authors_text":"Deyi Xiong, Hongfei Xu, Josef van Genabith, Qiuhui Liu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-09-30T13:18:57Z","title":"Learning Hard Retrieval Decoder Attention for Transformers"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2009.14658","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d21093340a65c13a5a911505108b99525a59992b1fd3c493bb59f27b35dc17ef","target":"record","created_at":"2026-07-05T03:13:00Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"a682c6b6887eb4619e1404cea25876087d5b2fe57b2311b198d4f050e464fd08","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-09-30T13:18:57Z","title_canon_sha256":"b62b720952f6ac31d0deacae62bc027f1ec48eeddd695acd2119524827ef8f36"},"schema_version":"1.0","source":{"id":"2009.14658","kind":"arxiv","version":2}},"canonical_sha256":"3090e85dadc8516a6341d3f7c4bec310e20d0db6d438046ee81a40cc037c90c1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3090e85dadc8516a6341d3f7c4bec310e20d0db6d438046ee81a40cc037c90c1","first_computed_at":"2026-07-05T03:13:00.457723Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T03:13:00.457723Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"KISPlFChe7m/wXDVyiwWjfzrB/PKQWQIfNwvsFy+hpRBiw+VQ1mMrNRW9OLODXFcf4rQ8UEjiEmETMHa06SaCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T03:13:00.458079Z","signed_message":"canonical_sha256_bytes"},"source_id":"2009.14658","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d21093340a65c13a5a911505108b99525a59992b1fd3c493bb59f27b35dc17ef","sha256:457d953e0cdc16373200678af06c0704441028ca975ef53fab8c3dd211b18822"],"state_sha256":"f29b5e2e283e543220160a94e768da42527ff197865b08df972cd2cf08bca2a9"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LJMNgXp8TmccTOhvOS/Nu5vIIVDKWrwieshJtcJph1go/ce4fWhO4S7LgBrTZ6f31eqi/IZKKGwuTh3IGW0hCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-08T13:57:54.617959Z","bundle_sha256":"838a228f3e80c9cc31a2ff4e608325cfcd13cd58da590c5c1e185d99ea125ebf"}}