{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DREQDZOTRSM7XBJQ5DYFXFCYY7","short_pith_number":"pith:DREQDZOT","schema_version":"1.0","canonical_sha256":"1c4901e5d38c99fb8530e8f05b9458c7e47be940b7d1c6f42c8c89305b364847","source":{"kind":"arxiv","id":"2408.15710","version":2},"attestation_state":"computed","paper":{"title":"Conan-embedding: General Text Embedding with More and Better Negative Samples","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Shiyu Li, Shizhe Chen, Xi Chen, Yang Tang","submitted_at":"2024-08-28T11:18:06Z","abstract_excerpt":"With the growing popularity of RAG, the capabilities of embedding models are gaining increasing attention. Embedding models are primarily trained through contrastive loss learning, with negative examples being a key component. Previous work has proposed various hard negative mining strategies, but these strategies are typically employed as preprocessing steps. In this paper, we propose the conan-embedding model, which maximizes the utilization of more and higher-quality negative examples. Specifically, since the model's ability to handle preprocessed negative examples evolves during training, "},"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":"2408.15710","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-28T11:18:06Z","cross_cats_sorted":[],"title_canon_sha256":"22520668e9dc00d15ec655bf76d8ac9b91b845b2f77baa01f8a91e2e6d2099df","abstract_canon_sha256":"2ae66844fa05d9a7ac5d854c85bc30f750d01bdec699c10254eb67f96c55271a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:00:39.130679Z","signature_b64":"HFKUrU0YiMK9L9rtwoCm5BVYVv0b6FnB5sMlE6DaEjeH1fV+3dUomsTDo/55LxdZ7p2IwJ1ZQEU2zhRzD5GUCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1c4901e5d38c99fb8530e8f05b9458c7e47be940b7d1c6f42c8c89305b364847","last_reissued_at":"2026-07-05T09:00:39.130241Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:00:39.130241Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Conan-embedding: General Text Embedding with More and Better Negative Samples","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Shiyu Li, Shizhe Chen, Xi Chen, Yang Tang","submitted_at":"2024-08-28T11:18:06Z","abstract_excerpt":"With the growing popularity of RAG, the capabilities of embedding models are gaining increasing attention. Embedding models are primarily trained through contrastive loss learning, with negative examples being a key component. Previous work has proposed various hard negative mining strategies, but these strategies are typically employed as preprocessing steps. In this paper, we propose the conan-embedding model, which maximizes the utilization of more and higher-quality negative examples. Specifically, since the model's ability to handle preprocessed negative examples evolves during training, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.15710","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/2408.15710/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":"2408.15710","created_at":"2026-07-05T09:00:39.130291+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.15710v2","created_at":"2026-07-05T09:00:39.130291+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.15710","created_at":"2026-07-05T09:00:39.130291+00:00"},{"alias_kind":"pith_short_12","alias_value":"DREQDZOTRSM7","created_at":"2026-07-05T09:00:39.130291+00:00"},{"alias_kind":"pith_short_16","alias_value":"DREQDZOTRSM7XBJQ","created_at":"2026-07-05T09:00:39.130291+00:00"},{"alias_kind":"pith_short_8","alias_value":"DREQDZOT","created_at":"2026-07-05T09:00:39.130291+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2510.01801","citing_title":"Detecting LLM-Generated Spam Reviews by Integrating Language Model Embeddings and Graph Neural Network","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04976","citing_title":"Tencent Advertising Algorithm Challenge 2025: All-Modality Generative Recommendation","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10937","citing_title":"Benchmarking and Enabling Efficient Chinese Medical Retrieval via Asymmetric Encoders","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7","json":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7.json","graph_json":"https://pith.science/api/pith-number/DREQDZOTRSM7XBJQ5DYFXFCYY7/graph.json","events_json":"https://pith.science/api/pith-number/DREQDZOTRSM7XBJQ5DYFXFCYY7/events.json","paper":"https://pith.science/paper/DREQDZOT"},"agent_actions":{"view_html":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7","download_json":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7.json","view_paper":"https://pith.science/paper/DREQDZOT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.15710&json=true","fetch_graph":"https://pith.science/api/pith-number/DREQDZOTRSM7XBJQ5DYFXFCYY7/graph.json","fetch_events":"https://pith.science/api/pith-number/DREQDZOTRSM7XBJQ5DYFXFCYY7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7/action/storage_attestation","attest_author":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7/action/author_attestation","sign_citation":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7/action/citation_signature","submit_replication":"https://pith.science/pith/DREQDZOTRSM7XBJQ5DYFXFCYY7/action/replication_record"}},"created_at":"2026-07-05T09:00:39.130291+00:00","updated_at":"2026-07-05T09:00:39.130291+00:00"}