{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:VMPRIPGKZT65TBFPWQAUAUYYSH","short_pith_number":"pith:VMPRIPGK","schema_version":"1.0","canonical_sha256":"ab1f143ccaccfdd984afb40140531891ca1bd0fa375532e4e9e86f5c8c196dd9","source":{"kind":"arxiv","id":"2308.11466","version":2},"attestation_state":"computed","paper":{"title":"SONAR: Sentence-Level Multimodal and Language-Agnostic Representations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Beno\\^it Sagot, Holger Schwenk, Paul-Ambroise Duquenne","submitted_at":"2023-08-22T14:25:15Z","abstract_excerpt":"We introduce SONAR, a new multilingual and multimodal fixed-size sentence embedding space. Our single text encoder, covering 200 languages, substantially outperforms existing sentence embeddings such as LASER3 and LabSE on the xsim and xsim++ multilingual similarity search tasks. Speech segments can be embedded in the same SONAR embedding space using language-specific speech encoders trained in a teacher-student setting on speech transcription data. Our encoders outperform existing speech encoders on similarity search tasks. We also provide a text decoder for 200 languages, which allows us to "},"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":"2308.11466","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-08-22T14:25:15Z","cross_cats_sorted":[],"title_canon_sha256":"ebc295aa55cb6733847a6ff72f435e5c25c3e9cd3193d8ff002cd7a21d05f2ba","abstract_canon_sha256":"b846c36ac8a6df73497f6086991b02095a97813cb85b617d55165b7693f11d6c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:43:52.095035Z","signature_b64":"Kw5v77AHsQUl+P2l1i9peUHeWEeOmGtGEY3hJVS2Uim1iEz/oblrer/4pd2skmhLUweRfsbLC8OelSfcDNR9Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ab1f143ccaccfdd984afb40140531891ca1bd0fa375532e4e9e86f5c8c196dd9","last_reissued_at":"2026-07-05T06:43:52.094532Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:43:52.094532Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SONAR: Sentence-Level Multimodal and Language-Agnostic Representations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Beno\\^it Sagot, Holger Schwenk, Paul-Ambroise Duquenne","submitted_at":"2023-08-22T14:25:15Z","abstract_excerpt":"We introduce SONAR, a new multilingual and multimodal fixed-size sentence embedding space. Our single text encoder, covering 200 languages, substantially outperforms existing sentence embeddings such as LASER3 and LabSE on the xsim and xsim++ multilingual similarity search tasks. Speech segments can be embedded in the same SONAR embedding space using language-specific speech encoders trained in a teacher-student setting on speech transcription data. Our encoders outperform existing speech encoders on similarity search tasks. We also provide a text decoder for 200 languages, which allows us to "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.11466","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/2308.11466/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":"2308.11466","created_at":"2026-07-05T06:43:52.094588+00:00"},{"alias_kind":"arxiv_version","alias_value":"2308.11466v2","created_at":"2026-07-05T06:43:52.094588+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.11466","created_at":"2026-07-05T06:43:52.094588+00:00"},{"alias_kind":"pith_short_12","alias_value":"VMPRIPGKZT65","created_at":"2026-07-05T06:43:52.094588+00:00"},{"alias_kind":"pith_short_16","alias_value":"VMPRIPGKZT65TBFP","created_at":"2026-07-05T06:43:52.094588+00:00"},{"alias_kind":"pith_short_8","alias_value":"VMPRIPGK","created_at":"2026-07-05T06:43:52.094588+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.18109","citing_title":"FLiP: Towards understanding and interpreting multimodal multilingual sentence embeddings","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2606.17967","citing_title":"Learning task-specific subspaces via interventional post-training of speech foundation models","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08748","citing_title":"HydraQE: OSU's Submission for the IWSLT 2026 Speech Translation Metrics Shared Task","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27378","citing_title":"Formalizing Latent Thoughts: Four Axioms of Thought Representation in LLMs","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31411","citing_title":"Linguistic Bias Mitigation for Spoofing Detection via Gradient Reversal and A Variational Information Bottleneck","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19130","citing_title":"EgoBabyVLM: Benchmarking Cross-Modal Learning from Naturalistic Egocentric Video Data","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09476","citing_title":"Align and Shine: Building High-Quality Sentence-Aligned Corpora for Multilingual Text Simplification","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05103","citing_title":"Text Corpora as Concept Fields: Black-Box Hallucination and Novelty Measurement","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2401.08281","citing_title":"The Faiss library","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05103","citing_title":"Text Corpora as Concept Fields: Black-Box Hallucination and Novelty Measurement","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21786","citing_title":"From Codebooks to VLMs: Evaluating Automated Visual Discourse Analysis for Climate Change on Social Media","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18109","citing_title":"FLiP: Towards understanding and interpreting multimodal multilingual sentence embeddings","ref_index":30,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH","json":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH.json","graph_json":"https://pith.science/api/pith-number/VMPRIPGKZT65TBFPWQAUAUYYSH/graph.json","events_json":"https://pith.science/api/pith-number/VMPRIPGKZT65TBFPWQAUAUYYSH/events.json","paper":"https://pith.science/paper/VMPRIPGK"},"agent_actions":{"view_html":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH","download_json":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH.json","view_paper":"https://pith.science/paper/VMPRIPGK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2308.11466&json=true","fetch_graph":"https://pith.science/api/pith-number/VMPRIPGKZT65TBFPWQAUAUYYSH/graph.json","fetch_events":"https://pith.science/api/pith-number/VMPRIPGKZT65TBFPWQAUAUYYSH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH/action/storage_attestation","attest_author":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH/action/author_attestation","sign_citation":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH/action/citation_signature","submit_replication":"https://pith.science/pith/VMPRIPGKZT65TBFPWQAUAUYYSH/action/replication_record"}},"created_at":"2026-07-05T06:43:52.094588+00:00","updated_at":"2026-07-05T06:43:52.094588+00:00"}