{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:CT75AOB75B6OT2QRDC25YJM6HE","short_pith_number":"pith:CT75AOB7","schema_version":"1.0","canonical_sha256":"14ffd0383fe87ce9ea1118b5dc259e392d7d0b4900b273a483411033a7d8bfce","source":{"kind":"arxiv","id":"2409.10173","version":3},"attestation_state":"computed","paper":{"title":"jina-embeddings-v3: Multilingual Embeddings With Task LoRA","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.IR"],"primary_cat":"cs.CL","authors_text":"Andreas Koukounas, Bo Wang, Feng Wang, Georgios Mastrapas, Han Xiao, Isabelle Mohr, Markus Krimmel, Michael G\\\"unther, Mohammad Kalim Akram, Nan Wang, Saba Sturua","submitted_at":"2024-09-16T11:10:29Z","abstract_excerpt":"We introduce jina-embeddings-v3, a novel text embedding model with 570 million parameters, achieves state-of-the-art performance on multilingual data and long-context retrieval tasks, supporting context lengths of up to 8192 tokens. The model includes a set of task-specific Low-Rank Adaptation (LoRA) adapters to generate high-quality embeddings for query-document retrieval, clustering, classification, and text matching. Evaluation on the MTEB benchmark shows that jina-embeddings-v3 outperforms the latest proprietary embeddings from OpenAI and Cohere on English tasks, while achieving superior p"},"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":"2409.10173","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-16T11:10:29Z","cross_cats_sorted":["cs.AI","cs.IR"],"title_canon_sha256":"ef1d21cfca087d56b118b5eed3f0fc6b90aac9983cbdc71382620628e25963d4","abstract_canon_sha256":"3b16c2dd42dfa461b529e9a213f7c0bec74ac67c1d4e2ebdebdf0c01d90dbfff"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:09:07.136523Z","signature_b64":"bw+9+9jJqybsRANeo6tVkT8Jy63OQ88605HmLkT9a8KUwx92yD6/tOWbzuZ9m77alWihYSKWv5s82Ucm2+i6CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"14ffd0383fe87ce9ea1118b5dc259e392d7d0b4900b273a483411033a7d8bfce","last_reissued_at":"2026-07-05T09:09:07.135999Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:09:07.135999Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"jina-embeddings-v3: Multilingual Embeddings With Task LoRA","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.IR"],"primary_cat":"cs.CL","authors_text":"Andreas Koukounas, Bo Wang, Feng Wang, Georgios Mastrapas, Han Xiao, Isabelle Mohr, Markus Krimmel, Michael G\\\"unther, Mohammad Kalim Akram, Nan Wang, Saba Sturua","submitted_at":"2024-09-16T11:10:29Z","abstract_excerpt":"We introduce jina-embeddings-v3, a novel text embedding model with 570 million parameters, achieves state-of-the-art performance on multilingual data and long-context retrieval tasks, supporting context lengths of up to 8192 tokens. The model includes a set of task-specific Low-Rank Adaptation (LoRA) adapters to generate high-quality embeddings for query-document retrieval, clustering, classification, and text matching. Evaluation on the MTEB benchmark shows that jina-embeddings-v3 outperforms the latest proprietary embeddings from OpenAI and Cohere on English tasks, while achieving superior p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.10173","kind":"arxiv","version":3},"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/2409.10173/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":"2409.10173","created_at":"2026-07-05T09:09:07.136070+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.10173v3","created_at":"2026-07-05T09:09:07.136070+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.10173","created_at":"2026-07-05T09:09:07.136070+00:00"},{"alias_kind":"pith_short_12","alias_value":"CT75AOB75B6O","created_at":"2026-07-05T09:09:07.136070+00:00"},{"alias_kind":"pith_short_16","alias_value":"CT75AOB75B6OT2QR","created_at":"2026-07-05T09:09:07.136070+00:00"},{"alias_kind":"pith_short_8","alias_value":"CT75AOB7","created_at":"2026-07-05T09:09:07.136070+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":27,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22807","citing_title":"KaLM-Reranker-V1: Fast but Not Late Interaction for Compressed Document Reranking","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13537","citing_title":"When Does Mixing Help? Analyzing Query Embedding Interpolation in Multilingual Dense Retrieval","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01561","citing_title":"S-SPPO: Semantic-Calibrated Self-Play Preference Optimization","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31142","citing_title":"On the Robustness of Multilingual Text Embedding Rankings Across Learning Tasks, Languages, and Benchmark Datasets","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30682","citing_title":"ALM2Vec: Learning Audio Embeddings for Universal Audio Retrieval with Large Audio-Language Models","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16608","citing_title":"To MRL or not to MRL: Text Embeddings are Robust to Truncation Without Matryoshka Learning, Except In Heavy Truncation Scenarios","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24079","citing_title":"TRACER: A Semantic-Aware Framework for Fine-Grained Contamination Detection in Code LLMs","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27345","citing_title":"MATCHA: Matching Text via Contrastive Semantic Alignment","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30120","citing_title":"No More K-means: Single-Stage Sparse Coding for Efficient Multi-Vector Retrieval","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2504.14707","citing_title":"FLAME: A New Dataset on FLemish Accounts of Momentary Experiences","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13415","citing_title":"KIT-TIP-NLP at MultiPride: Continual Learning with Multilingual Foundation Model","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16608","citing_title":"To MRL or not to MRL: Text Embeddings are Robust to Truncation Without Matryoshka Learning, Except In Heavy Truncation Scenarios","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18153","citing_title":"Three Heads Are Better Than One: A Multi-perspective Reasoning Framework for Enhanced Vulnerability Detection","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2508.01959","citing_title":"SitEmb-v1.5: Improved Context-Aware Dense Retrieval for Semantic Association and Long Story Comprehension","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2601.03790","citing_title":"NeoAMT: Neologism-Aware Agentic Machine Translation with Reinforcement Learning","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2603.01059","citing_title":"GroupGPT: A Token-efficient and Privacy-preserving Agentic Framework for Multi-User Chat Assistant","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2509.20354","citing_title":"EmbeddingGemma: Powerful and Lightweight Text Representations","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12714","citing_title":"Layer-wise Representation Dynamics: An Empirical Investigation Across Embedders and Base LLMs","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13415","citing_title":"KIT-TIP-NLP at MultiPride: Continual Learning with Multilingual Foundation Model","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04962","citing_title":"TabEmbed: Benchmarking and Learning Generalist Embeddings for Tabular Understanding","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00297","citing_title":"Trident: Improving Malware Detection with LLMs and Behavioral Features","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20720","citing_title":"COMPASS: COntinual Multilingual PEFT with Adaptive Semantic Sampling","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07809","citing_title":"PolicyLong: Towards On-Policy Context Extension","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07249","citing_title":"MLAIRE: Multilingual Language-Aware Information Retrieval Evaluation Protocal","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17738","citing_title":"Mira-Embeddings-V1: Domain-Adapted Semantic Reranking for Recruitment via LLM-Synthesized Data","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE","json":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE.json","graph_json":"https://pith.science/api/pith-number/CT75AOB75B6OT2QRDC25YJM6HE/graph.json","events_json":"https://pith.science/api/pith-number/CT75AOB75B6OT2QRDC25YJM6HE/events.json","paper":"https://pith.science/paper/CT75AOB7"},"agent_actions":{"view_html":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE","download_json":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE.json","view_paper":"https://pith.science/paper/CT75AOB7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.10173&json=true","fetch_graph":"https://pith.science/api/pith-number/CT75AOB75B6OT2QRDC25YJM6HE/graph.json","fetch_events":"https://pith.science/api/pith-number/CT75AOB75B6OT2QRDC25YJM6HE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE/action/storage_attestation","attest_author":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE/action/author_attestation","sign_citation":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE/action/citation_signature","submit_replication":"https://pith.science/pith/CT75AOB75B6OT2QRDC25YJM6HE/action/replication_record"}},"created_at":"2026-07-05T09:09:07.136070+00:00","updated_at":"2026-07-05T09:09:07.136070+00:00"}