{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:EVH2VCU6DHZ3RXW4BTMUCOL5BA","short_pith_number":"pith:EVH2VCU6","schema_version":"1.0","canonical_sha256":"254faa8a9e19f3b8dedc0cd941397d083fcd62bd93b3f6ea6c277cd16a482742","source":{"kind":"arxiv","id":"1902.08564","version":2},"attestation_state":"computed","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"},"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":"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"},"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"},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"1902.08564","created_at":"2026-05-17T23:43:14.671455+00:00"},{"alias_kind":"arxiv_version","alias_value":"1902.08564v2","created_at":"2026-05-17T23:43:14.671455+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1902.08564","created_at":"2026-05-17T23:43:14.671455+00:00"},{"alias_kind":"pith_short_12","alias_value":"EVH2VCU6DHZ3","created_at":"2026-05-18T12:33:15.570797+00:00"},{"alias_kind":"pith_short_16","alias_value":"EVH2VCU6DHZ3RXW4","created_at":"2026-05-18T12:33:15.570797+00:00"},{"alias_kind":"pith_short_8","alias_value":"EVH2VCU6","created_at":"2026-05-18T12:33:15.570797+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/EVH2VCU6DHZ3RXW4BTMUCOL5BA","json":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA.json","graph_json":"https://pith.science/api/pith-number/EVH2VCU6DHZ3RXW4BTMUCOL5BA/graph.json","events_json":"https://pith.science/api/pith-number/EVH2VCU6DHZ3RXW4BTMUCOL5BA/events.json","paper":"https://pith.science/paper/EVH2VCU6"},"agent_actions":{"view_html":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA","download_json":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA.json","view_paper":"https://pith.science/paper/EVH2VCU6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1902.08564&json=true","fetch_graph":"https://pith.science/api/pith-number/EVH2VCU6DHZ3RXW4BTMUCOL5BA/graph.json","fetch_events":"https://pith.science/api/pith-number/EVH2VCU6DHZ3RXW4BTMUCOL5BA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/action/storage_attestation","attest_author":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/action/author_attestation","sign_citation":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/action/citation_signature","submit_replication":"https://pith.science/pith/EVH2VCU6DHZ3RXW4BTMUCOL5BA/action/replication_record"}},"created_at":"2026-05-17T23:43:14.671455+00:00","updated_at":"2026-05-17T23:43:14.671455+00:00"}