{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:EVH2VCU6DHZ3RXW4BTMUCOL5BA","short_pith_number":"pith:EVH2VCU6","canonical_record":{"source":{"id":"1902.08564","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-02-22T17:11:03Z","cross_cats_sorted":[],"title_canon_sha256":"8bc52384497709ae6d75d7db1d56b3580b947d6a531fc56ad014627aaf020213","abstract_canon_sha256":"d62d3af885a1729c8c88e1884641e6375bc66dac1e471f5bad885a507cb83185"},"schema_version":"1.0"},"canonical_sha256":"254faa8a9e19f3b8dedc0cd941397d083fcd62bd93b3f6ea6c277cd16a482742","source":{"kind":"arxiv","id":"1902.08564","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1902.08564","created_at":"2026-05-17T23:43:14Z"},{"alias_kind":"arxiv_version","alias_value":"1902.08564v2","created_at":"2026-05-17T23:43:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1902.08564","created_at":"2026-05-17T23:43:14Z"},{"alias_kind":"pith_short_12","alias_value":"EVH2VCU6DHZ3","created_at":"2026-05-18T12:33:15Z"},{"alias_kind":"pith_short_16","alias_value":"EVH2VCU6DHZ3RXW4","created_at":"2026-05-18T12:33:15Z"},{"alias_kind":"pith_short_8","alias_value":"EVH2VCU6","created_at":"2026-05-18T12:33:15Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:EVH2VCU6DHZ3RXW4BTMUCOL5BA","target":"record","payload":{"canonical_record":{"source":{"id":"1902.08564","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-02-22T17:11:03Z","cross_cats_sorted":[],"title_canon_sha256":"8bc52384497709ae6d75d7db1d56b3580b947d6a531fc56ad014627aaf020213","abstract_canon_sha256":"d62d3af885a1729c8c88e1884641e6375bc66dac1e471f5bad885a507cb83185"},"schema_version":"1.0"},"canonical_sha256":"254faa8a9e19f3b8dedc0cd941397d083fcd62bd93b3f6ea6c277cd16a482742","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:43:14.671958Z","signature_b64":"O+plYRv+kmJbnCLX0D44MvliqvhoJJNrrA4nGRa+T4Oh/ZNsy4XOg6OZEATzRKpIojChttw6wqjqq/IUwFCsDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"254faa8a9e19f3b8dedc0cd941397d083fcd62bd93b3f6ea6c277cd16a482742","last_reissued_at":"2026-05-17T23:43:14.671368Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:43:14.671368Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1902.08564","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":"2026-05-17T23:43:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gAx1wLk7VGFmS6fHeTIzoNClSVXfBbkNidzhK9shzLsCaKzdRCFwCjFUIp9H9mjjiAJ1PyF3XXAt5oD/ZmASAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T10:04:47.177963Z"},"content_sha256":"ad9d273d450a23da27c2f61b3f84684cef0164b9b3c436614532a990dfe57117","schema_version":"1.0","event_id":"sha256:ad9d273d450a23da27c2f61b3f84684cef0164b9b3c436614532a990dfe57117"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:EVH2VCU6DHZ3RXW4BTMUCOL5BA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Improving Multilingual Sentence Embedding using Bi-directional Dual Encoder with Additive Margin Softmax","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Brian Strope, Daniel Cer, Gustavo Hernandez Abrego, Mandy Guo, Qinlan Shen, Ray Kurzweil, Steve Yuan, Yinfei Yang, Yun-hsuan Sung","submitted_at":"2019-02-22T17:11:03Z","abstract_excerpt":"In this paper, we present an approach to learn multilingual sentence embeddings using a bi-directional dual-encoder with additive margin softmax. The embeddings are able to achieve state-of-the-art results on the United Nations (UN) parallel corpus retrieval task. In all the languages tested, the system achieves P@1 of 86% or higher. We use pairs retrieved by our approach to train NMT models that achieve similar performance to models trained on gold pairs. We explore simple document-level embeddings constructed by averaging our sentence embeddings. On the UN document-level retrieval task, docu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1902.08564","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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:43:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YYitIoXEpo/O5mDqiz5/pttJshlRHMqRHGBYBmrKXiOPW5eR4vNHg4CJ2I+DDr/+aR41j8Q99EYWqKTQziwXCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T10:04:47.178631Z"},"content_sha256":"e9792511c5ac3d4a2f5c2d942b31a3873abc9fc6dc5cf73ab563f1ffcbd80be3","schema_version":"1.0","event_id":"sha256:e9792511c5ac3d4a2f5c2d942b31a3873abc9fc6dc5cf73ab563f1ffcbd80be3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/bundle.json","state_url":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/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":"2026-05-30T10:04:47Z","links":{"resolver":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA","bundle":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/bundle.json","state":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:EVH2VCU6DHZ3RXW4BTMUCOL5BA","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":"d62d3af885a1729c8c88e1884641e6375bc66dac1e471f5bad885a507cb83185","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-02-22T17:11:03Z","title_canon_sha256":"8bc52384497709ae6d75d7db1d56b3580b947d6a531fc56ad014627aaf020213"},"schema_version":"1.0","source":{"id":"1902.08564","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1902.08564","created_at":"2026-05-17T23:43:14Z"},{"alias_kind":"arxiv_version","alias_value":"1902.08564v2","created_at":"2026-05-17T23:43:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1902.08564","created_at":"2026-05-17T23:43:14Z"},{"alias_kind":"pith_short_12","alias_value":"EVH2VCU6DHZ3","created_at":"2026-05-18T12:33:15Z"},{"alias_kind":"pith_short_16","alias_value":"EVH2VCU6DHZ3RXW4","created_at":"2026-05-18T12:33:15Z"},{"alias_kind":"pith_short_8","alias_value":"EVH2VCU6","created_at":"2026-05-18T12:33:15Z"}],"graph_snapshots":[{"event_id":"sha256:e9792511c5ac3d4a2f5c2d942b31a3873abc9fc6dc5cf73ab563f1ffcbd80be3","target":"graph","created_at":"2026-05-17T23:43:14Z","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"},"paper":{"abstract_excerpt":"In this paper, we present an approach to learn multilingual sentence embeddings using a bi-directional dual-encoder with additive margin softmax. The embeddings are able to achieve state-of-the-art results on the United Nations (UN) parallel corpus retrieval task. In all the languages tested, the system achieves P@1 of 86% or higher. We use pairs retrieved by our approach to train NMT models that achieve similar performance to models trained on gold pairs. We explore simple document-level embeddings constructed by averaging our sentence embeddings. On the UN document-level retrieval task, docu","authors_text":"Brian Strope, Daniel Cer, Gustavo Hernandez Abrego, Mandy Guo, Qinlan Shen, Ray Kurzweil, Steve Yuan, Yinfei Yang, Yun-hsuan Sung","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-02-22T17:11:03Z","title":"Improving Multilingual Sentence Embedding using Bi-directional Dual Encoder with Additive Margin Softmax"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1902.08564","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:ad9d273d450a23da27c2f61b3f84684cef0164b9b3c436614532a990dfe57117","target":"record","created_at":"2026-05-17T23:43:14Z","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":"d62d3af885a1729c8c88e1884641e6375bc66dac1e471f5bad885a507cb83185","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-02-22T17:11:03Z","title_canon_sha256":"8bc52384497709ae6d75d7db1d56b3580b947d6a531fc56ad014627aaf020213"},"schema_version":"1.0","source":{"id":"1902.08564","kind":"arxiv","version":2}},"canonical_sha256":"254faa8a9e19f3b8dedc0cd941397d083fcd62bd93b3f6ea6c277cd16a482742","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"254faa8a9e19f3b8dedc0cd941397d083fcd62bd93b3f6ea6c277cd16a482742","first_computed_at":"2026-05-17T23:43:14.671368Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:43:14.671368Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"O+plYRv+kmJbnCLX0D44MvliqvhoJJNrrA4nGRa+T4Oh/ZNsy4XOg6OZEATzRKpIojChttw6wqjqq/IUwFCsDg==","signature_status":"signed_v1","signed_at":"2026-05-17T23:43:14.671958Z","signed_message":"canonical_sha256_bytes"},"source_id":"1902.08564","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ad9d273d450a23da27c2f61b3f84684cef0164b9b3c436614532a990dfe57117","sha256:e9792511c5ac3d4a2f5c2d942b31a3873abc9fc6dc5cf73ab563f1ffcbd80be3"],"state_sha256":"2f5aa226dea7f02d17787a4dd3db57c67ad7f0c1ef42d5d727d72173a0c0e080"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"L9jtVbc6OWYqhWE5LJ0quCk+TzpccDwj76CNW11iFknra+Axa+XRQrtW1x8wws5QHlgeMXDHXyflu572KdwDBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T10:04:47.181132Z","bundle_sha256":"4e7ce927b05932513230b695b54353cee5fe1ccfc73754f8510424611afdd980"}}