{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DBHOAQEYUAU7L7KCXWHPMLDY3M","short_pith_number":"pith:DBHOAQEY","schema_version":"1.0","canonical_sha256":"184ee04098a029f5fd42bd8ef62c78db11ab4a6dca38137287347cc9874dcec3","source":{"kind":"arxiv","id":"2501.01028","version":4},"attestation_state":"computed","paper":{"title":"KaLM-Embedding: Superior Training Data Brings A Stronger Embedding Model","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Baotian Hu, Dongfang Li, Haofen Wang, Jun Yu, Min Zhang, Qian Chen, Shaolin Ye, Xinping Zhao, Xinshuo Hu, Xinyuan Wei, Zetian Sun, Zhenyu Liu, Zifei Shan","submitted_at":"2025-01-02T03:17:51Z","abstract_excerpt":"As retrieval-augmented generation prevails in large language models, embedding models are becoming increasingly crucial. Despite the growing number of general embedding models, prior work often overlooks the critical role of training data quality. In this work, we introduce KaLM-Embedding, a general multilingual embedding model that leverages a large quantity of cleaner, more diverse, and domain-specific training data. Our model has been trained with key techniques proven to enhance performance: (1) persona-based synthetic data to create diversified examples distilled from LLMs, (2) ranking co"},"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":"2501.01028","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-02T03:17:51Z","cross_cats_sorted":[],"title_canon_sha256":"00f89c23d528466a5a821077081df95d830c4d0ed5389e7dc66e1fbf73596af6","abstract_canon_sha256":"6d1fe9ca1aaadd9397c1cda34d650c42b6f6139457221d4dffe9e4719bfa0f5c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:01:18.997964Z","signature_b64":"CrfHc8Co481Mp/45TctQ9ECGgZofH1jUntjOwRmKjMII4xiBd7QV58VfSNQBQ3jT7ISFiQCltpPKh3vWaj5qCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"184ee04098a029f5fd42bd8ef62c78db11ab4a6dca38137287347cc9874dcec3","last_reissued_at":"2026-07-05T10:01:18.997496Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:01:18.997496Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"KaLM-Embedding: Superior Training Data Brings A Stronger Embedding Model","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Baotian Hu, Dongfang Li, Haofen Wang, Jun Yu, Min Zhang, Qian Chen, Shaolin Ye, Xinping Zhao, Xinshuo Hu, Xinyuan Wei, Zetian Sun, Zhenyu Liu, Zifei Shan","submitted_at":"2025-01-02T03:17:51Z","abstract_excerpt":"As retrieval-augmented generation prevails in large language models, embedding models are becoming increasingly crucial. Despite the growing number of general embedding models, prior work often overlooks the critical role of training data quality. In this work, we introduce KaLM-Embedding, a general multilingual embedding model that leverages a large quantity of cleaner, more diverse, and domain-specific training data. Our model has been trained with key techniques proven to enhance performance: (1) persona-based synthetic data to create diversified examples distilled from LLMs, (2) ranking co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.01028","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/2501.01028/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":"2501.01028","created_at":"2026-07-05T10:01:18.997555+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.01028v4","created_at":"2026-07-05T10:01:18.997555+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.01028","created_at":"2026-07-05T10:01:18.997555+00:00"},{"alias_kind":"pith_short_12","alias_value":"DBHOAQEYUAU7","created_at":"2026-07-05T10:01:18.997555+00:00"},{"alias_kind":"pith_short_16","alias_value":"DBHOAQEYUAU7L7KC","created_at":"2026-07-05T10:01:18.997555+00:00"},{"alias_kind":"pith_short_8","alias_value":"DBHOAQEY","created_at":"2026-07-05T10:01:18.997555+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"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":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10254","citing_title":"RealMath-Eval: Why SOTA Judges Struggle with Real Human Reasoning","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03027","citing_title":"SEA-Embedding: Open and Reproducible Text Embeddings for Southeast Asia","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24297","citing_title":"Benchmarking Patent Embeddings: A Multi-Task Evaluation of 22 Models Across Retrieval, Classification, and Clustering","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27810","citing_title":"LRanker: LLM Ranker for Massive Candidates","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22202","citing_title":"Structure Retention in Embedding Spaces as a Predictor of Benchmark Performance","ref_index":98,"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":1,"is_internal_anchor":false},{"citing_arxiv_id":"2603.12572","citing_title":"LMEB: Long-horizon Memory Embedding Benchmark","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23734","citing_title":"Prism-Reranker: Beyond Relevance Scoring -- Jointly Producing Contributions and Evidence for Agentic Retrieval","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M","json":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M.json","graph_json":"https://pith.science/api/pith-number/DBHOAQEYUAU7L7KCXWHPMLDY3M/graph.json","events_json":"https://pith.science/api/pith-number/DBHOAQEYUAU7L7KCXWHPMLDY3M/events.json","paper":"https://pith.science/paper/DBHOAQEY"},"agent_actions":{"view_html":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M","download_json":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M.json","view_paper":"https://pith.science/paper/DBHOAQEY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.01028&json=true","fetch_graph":"https://pith.science/api/pith-number/DBHOAQEYUAU7L7KCXWHPMLDY3M/graph.json","fetch_events":"https://pith.science/api/pith-number/DBHOAQEYUAU7L7KCXWHPMLDY3M/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M/action/storage_attestation","attest_author":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M/action/author_attestation","sign_citation":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M/action/citation_signature","submit_replication":"https://pith.science/pith/DBHOAQEYUAU7L7KCXWHPMLDY3M/action/replication_record"}},"created_at":"2026-07-05T10:01:18.997555+00:00","updated_at":"2026-07-05T10:01:18.997555+00:00"}