{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:GEYPE2F54RO7ETGBDXPUQDIYHH","short_pith_number":"pith:GEYPE2F5","schema_version":"1.0","canonical_sha256":"3130f268bde45df24cc11ddf480d1839f161edf2e7bc1ca3f10af5b709f901c0","source":{"kind":"arxiv","id":"2007.00808","version":2},"attestation_state":"computed","paper":{"title":"Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.IR","authors_text":"Arnold Overwijk, Chenyan Xiong, Jialin Liu, Junaid Ahmed, Kwok-Fung Tang, Lee Xiong, Paul Bennett, Ye Li","submitted_at":"2020-07-01T23:15:56Z","abstract_excerpt":"Conducting text retrieval in a dense learned representation space has many intriguing advantages over sparse retrieval. Yet the effectiveness of dense retrieval (DR) often requires combination with sparse retrieval. In this paper, we identify that the main bottleneck is in the training mechanisms, where the negative instances used in training are not representative of the irrelevant documents in testing. This paper presents Approximate nearest neighbor Negative Contrastive Estimation (ANCE), a training mechanism that constructs negatives from an Approximate Nearest Neighbor (ANN) index of the "},"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":"2007.00808","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2020-07-01T23:15:56Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"555b005abbac6178ab01c17cd261815599964df2c06be23d81af176d10876441","abstract_canon_sha256":"1efd50254258e8be91a5b204910063254dfe8d9ef3322fd9e2089e69f057e85b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:44:49.370581Z","signature_b64":"U8tWTVro716MJlFkiaWEW2Cop1kRCYEvFQZOlngwvzV02o0ARWRk7IFLcBlEXonb0YJ+mS8nGfp8LdNvvWejCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3130f268bde45df24cc11ddf480d1839f161edf2e7bc1ca3f10af5b709f901c0","last_reissued_at":"2026-07-05T01:44:49.370100Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:44:49.370100Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.IR","authors_text":"Arnold Overwijk, Chenyan Xiong, Jialin Liu, Junaid Ahmed, Kwok-Fung Tang, Lee Xiong, Paul Bennett, Ye Li","submitted_at":"2020-07-01T23:15:56Z","abstract_excerpt":"Conducting text retrieval in a dense learned representation space has many intriguing advantages over sparse retrieval. Yet the effectiveness of dense retrieval (DR) often requires combination with sparse retrieval. In this paper, we identify that the main bottleneck is in the training mechanisms, where the negative instances used in training are not representative of the irrelevant documents in testing. This paper presents Approximate nearest neighbor Negative Contrastive Estimation (ANCE), a training mechanism that constructs negatives from an Approximate Nearest Neighbor (ANN) index of the "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2007.00808","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/2007.00808/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":"2007.00808","created_at":"2026-07-05T01:44:49.370157+00:00"},{"alias_kind":"arxiv_version","alias_value":"2007.00808v2","created_at":"2026-07-05T01:44:49.370157+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2007.00808","created_at":"2026-07-05T01:44:49.370157+00:00"},{"alias_kind":"pith_short_12","alias_value":"GEYPE2F54RO7","created_at":"2026-07-05T01:44:49.370157+00:00"},{"alias_kind":"pith_short_16","alias_value":"GEYPE2F54RO7ETGB","created_at":"2026-07-05T01:44:49.370157+00:00"},{"alias_kind":"pith_short_8","alias_value":"GEYPE2F5","created_at":"2026-07-05T01:44:49.370157+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":34,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.00004","citing_title":"Why Advanced Encoders Lag on Sparse Retrieval? The Answer and an Approach to Bridging Vocabulary Gaps","ref_index":60,"is_internal_anchor":true},{"citing_arxiv_id":"2606.25674","citing_title":"BitNet Text Embeddings","ref_index":71,"is_internal_anchor":false},{"citing_arxiv_id":"2606.24200","citing_title":"MMed-Bench-IR: A Heterogeneous Benchmark for Multilingual Medical Information Retrieval","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2606.18310","citing_title":"Conflict-Aware Retriever Editing for Knowledge Injection Attacks on LLM-Based RAG Systems","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13204","citing_title":"CoDeR: Local Constraint-Compatible Retrieval Beyond Semantic Similarity","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00012","citing_title":"PRA-RAG: Provably Robust Aggregation in Retrieval-Augmented Generation against Retrieval Corruption","ref_index":127,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07924","citing_title":"Decoupling Semantics and Logic: A Training-Free Coarse-to-Fine Pipeline for Video Retrieval-Augmented Generation","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03535","citing_title":"Can LLM Rerankers Predict Their Own Ranking Performance?","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00833","citing_title":"Generative Retrieval for Table Union Search","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01070","citing_title":"Test-Time Training for Zero-Resource Dense Retrieval Reranking","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26002","citing_title":"SemBridge: Language Transfer in Sparse Encoders via Multilingual Semantic Bridges","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28641","citing_title":"Subtraction Gets You More: Gap-Aware Retrieval for Multimodal Multi-Hop QA","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01482","citing_title":"Beyond Topical Similarity: Contrastive Evidence Retrieval with Interpretable Attention Alignment in RAG","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2412.14751","citing_title":"Query pipeline optimization for cancer patient question answering systems","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2504.02181","citing_title":"A Survey of Scaling in Large Language Model Reasoning","ref_index":233,"is_internal_anchor":false},{"citing_arxiv_id":"2504.21015","citing_title":"Don't Retrieve, Generate: Prompting LLMs for Synthetic Training Data in Dense Retrieval","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2507.08480","citing_title":"Improving Korean-English Cross-Lingual Retrieval: A Data-Centric Study of Language Composition and Model Merging","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2509.16621","citing_title":"The Role of Vocabularies in Learning Sparse Representations for Ranking","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2510.18333","citing_title":"Position: LLM Watermarking Should Align Stakeholders' Incentives for Practical Adoption","ref_index":59,"is_internal_anchor":false},{"citing_arxiv_id":"2410.05160","citing_title":"VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2208.03299","citing_title":"Atlas: Few-shot Learning with Retrieval Augmented Language Models","ref_index":259,"is_internal_anchor":false},{"citing_arxiv_id":"2004.04906","citing_title":"Dense Passage Retrieval for Open-Domain Question Answering","ref_index":109,"is_internal_anchor":false},{"citing_arxiv_id":"2201.10005","citing_title":"Text and Code Embeddings by Contrastive Pre-Training","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14169","citing_title":"BOOKMARKS: Efficient Active Storyline Memory for Role-playing","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12736","citing_title":"ConRetroBert: EMA Stabilized Dual Encoders for Template-Based Single-Step Retrosynthesis","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH","json":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH.json","graph_json":"https://pith.science/api/pith-number/GEYPE2F54RO7ETGBDXPUQDIYHH/graph.json","events_json":"https://pith.science/api/pith-number/GEYPE2F54RO7ETGBDXPUQDIYHH/events.json","paper":"https://pith.science/paper/GEYPE2F5"},"agent_actions":{"view_html":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH","download_json":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH.json","view_paper":"https://pith.science/paper/GEYPE2F5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2007.00808&json=true","fetch_graph":"https://pith.science/api/pith-number/GEYPE2F54RO7ETGBDXPUQDIYHH/graph.json","fetch_events":"https://pith.science/api/pith-number/GEYPE2F54RO7ETGBDXPUQDIYHH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH/action/storage_attestation","attest_author":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH/action/author_attestation","sign_citation":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH/action/citation_signature","submit_replication":"https://pith.science/pith/GEYPE2F54RO7ETGBDXPUQDIYHH/action/replication_record"}},"created_at":"2026-07-05T01:44:49.370157+00:00","updated_at":"2026-07-05T01:44:49.370157+00:00"}