{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:FE3TD4REDRODANF4VAMY3HCDWL","short_pith_number":"pith:FE3TD4RE","schema_version":"1.0","canonical_sha256":"293731f2241c5c3034bca8198d9c43b2ea7482fd28cbfdde07a326f2610f445b","source":{"kind":"arxiv","id":"2004.05160","version":4},"attestation_state":"computed","paper":{"title":"On the Language Neutrality of Pre-trained Multilingual Representations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alexander Fraser, Jind\\v{r}ich Libovick\\'y, Rudolf Rosa","submitted_at":"2020-04-09T19:50:32Z","abstract_excerpt":"Multilingual contextual embeddings, such as multilingual BERT and XLM-RoBERTa, have proved useful for many multi-lingual tasks. Previous work probed the cross-linguality of the representations indirectly using zero-shot transfer learning on morphological and syntactic tasks. We instead investigate the language-neutrality of multilingual contextual embeddings directly and with respect to lexical semantics. Our results show that contextual embeddings are more language-neutral and, in general, more informative than aligned static word-type embeddings, which are explicitly trained for language neu"},"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":"2004.05160","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-09T19:50:32Z","cross_cats_sorted":[],"title_canon_sha256":"358796471a8ea0e478cec747432ef03c41481d24a3b7b8464d78f8c39104e233","abstract_canon_sha256":"0eb6d16e6e609a97e6ea28d5cd613de06ebe4d8a1e3087512ac0020fc9c1a8bf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:39:12.788802Z","signature_b64":"Jm6Tuc97f2xDCvRJc/WG7ItNlP0FAPsYLIkpNzT0i2LWiDAeByX9TUL/74TcA0zUetKoGIhNjhuVKcFW8R/dDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"293731f2241c5c3034bca8198d9c43b2ea7482fd28cbfdde07a326f2610f445b","last_reissued_at":"2026-07-05T01:39:12.788373Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:39:12.788373Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On the Language Neutrality of Pre-trained Multilingual Representations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alexander Fraser, Jind\\v{r}ich Libovick\\'y, Rudolf Rosa","submitted_at":"2020-04-09T19:50:32Z","abstract_excerpt":"Multilingual contextual embeddings, such as multilingual BERT and XLM-RoBERTa, have proved useful for many multi-lingual tasks. Previous work probed the cross-linguality of the representations indirectly using zero-shot transfer learning on morphological and syntactic tasks. We instead investigate the language-neutrality of multilingual contextual embeddings directly and with respect to lexical semantics. Our results show that contextual embeddings are more language-neutral and, in general, more informative than aligned static word-type embeddings, which are explicitly trained for language neu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.05160","kind":"arxiv","version":4},"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/2004.05160/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":"2004.05160","created_at":"2026-07-05T01:39:12.788437+00:00"},{"alias_kind":"arxiv_version","alias_value":"2004.05160v4","created_at":"2026-07-05T01:39:12.788437+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.05160","created_at":"2026-07-05T01:39:12.788437+00:00"},{"alias_kind":"pith_short_12","alias_value":"FE3TD4REDROD","created_at":"2026-07-05T01:39:12.788437+00:00"},{"alias_kind":"pith_short_16","alias_value":"FE3TD4REDRODANF4","created_at":"2026-07-05T01:39:12.788437+00:00"},{"alias_kind":"pith_short_8","alias_value":"FE3TD4RE","created_at":"2026-07-05T01:39:12.788437+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.10627","citing_title":"Computational Lesions in Multilingual Language Models Separate Shared and Language-specific Brain Alignment","ref_index":82,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL","json":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL.json","graph_json":"https://pith.science/api/pith-number/FE3TD4REDRODANF4VAMY3HCDWL/graph.json","events_json":"https://pith.science/api/pith-number/FE3TD4REDRODANF4VAMY3HCDWL/events.json","paper":"https://pith.science/paper/FE3TD4RE"},"agent_actions":{"view_html":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL","download_json":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL.json","view_paper":"https://pith.science/paper/FE3TD4RE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2004.05160&json=true","fetch_graph":"https://pith.science/api/pith-number/FE3TD4REDRODANF4VAMY3HCDWL/graph.json","fetch_events":"https://pith.science/api/pith-number/FE3TD4REDRODANF4VAMY3HCDWL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL/action/storage_attestation","attest_author":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL/action/author_attestation","sign_citation":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL/action/citation_signature","submit_replication":"https://pith.science/pith/FE3TD4REDRODANF4VAMY3HCDWL/action/replication_record"}},"created_at":"2026-07-05T01:39:12.788437+00:00","updated_at":"2026-07-05T01:39:12.788437+00:00"}