mlsanityData to 7 Oct 2026

SimANS: Simple Ambiguous Negatives Sampling for Dense Text Retrieval

arXiv 2210.11773 ↗First posted 21 Oct 2022cs.CL cs.IRcited by 21

Not on the map. Why?

Abstract (arXiv)

Sampling proper negatives from a large document pool is vital to effectively train a dense retrieval model. However, existing negative sampling strategies suffer from the uninformative or false negative problem. In this work, we empirically show that according to the measured relevance scores, the negatives ranked around the positives are generally more informative and less likely to be false negatives. Intuitively, these negatives are not too hard (may be false negatives) or too easy (uninformative). They are the ambiguous negatives and need more attention during training. Thus, we propose a simple ambiguous negatives sampling method, SimANS, which incorporates a new sampling probability distribution to sample more ambiguous negatives. Extensive experiments on four public and one industry datasets show the effectiveness of our approach. We made the code and models publicly available in urlhttps://github.com/microsoft/SimXNS.

Key facts. Cited by 21 papers in the map (data through 7 Oct 2026). Builds on 39 of its 53 references that are in the map. First posted on arXiv 21 Oct 2022.

Builds on

39 of 53 references.

References · 39

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,049
  3. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,779
  4. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,895
  5. 2017Billion-scale similarity search with GPUscited by 2,477
  6. 2017Understanding Black-box Predictions via Influence Functionscited by 1,714
  7. 2016MS MARCO: A Human Generated MAchine Reading COmprehension Datasetcited by 1,591
  8. 2020Leveraging Passage Retrieval with Generative Models for Open Domain Question Answeringcited by 1,203
  9. 2020Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrievalcited by 846
  10. 2020Estimating Training Data Influence by Tracing Gradient Descentcited by 565
  11. 2020Dataset Cartography: Mapping and Diagnosing Datasets with Training Dynamicscited by 423
  12. 2020RocketQA: An Optimized Training Approach to Dense Passage Retrieval for Open-Domain Question Answeringcited by 410
27 more
  1. 2020AmbigQA: Answering Ambiguous Open-domain Questionscited by 385
  2. 2018Not All Samples Are Created Equal: Deep Learning with Importance Samplingcited by 323
  3. 2021Efficiently Teaching an Effective Dense Retriever with Balanced Topic Aware Samplingcited by 238
  4. 2019ERNIE 2.0: A Continual Pre-training Framework for Language Understandingcited by 230
  5. 2019Document Expansion by Query Predictioncited by 225
  6. 2020Sparse, Dense, and Attentional Representations for Text Retrievalcited by 213
  7. 2021Unsupervised Corpus Aware Language Model Pre-training for Dense Passage Retrievalcited by 203
  8. 2020Generation-Augmented Retrieval for Open-domain Question Answeringcited by 186
  9. 2021RocketQAv2: A Joint Training Method for Dense Passage Retrieval and Passage Re-rankingcited by 175
  10. 2021Optimizing Dense Retrieval Model Training with Hard Negativescited by 154
  11. 2021Scaling Deep Contrastive Learning Batch Size under Memory Limited Setupcited by 123
  12. 2021COIL: Revisit Exact Lexical Match in Information Retrieval with Contextualized Inverted Listcited by 113
  13. 2019Deeper Text Understanding for IR with Contextual Neural Language Modelingcited by 111
  14. 2021Adversarial Retriever-Ranker for dense text retrievalcited by 82
  15. 2021End-to-End Training of Neural Retrievers for Open-Domain Question Answeringcited by 75
  16. 2020Measuring Social Biases in Grounded Vision and Language Embeddingscited by 61
  17. 2022Debiased Contrastive Learning of Unsupervised Sentence Representationscited by 56
  18. 2021Domain-matched Pre-training Tasks for Dense Retrievalcited by 53
  19. 2021PAIR: Leveraging Passage-Centric Similarity Relation for Improving Dense Passage Retrievalcited by 50
  20. 2020RepBERT: Contextualized Text Embeddings for First-Stage Retrievalcited by 46
  21. 2021Learning to Retrieve Passages without Supervisioncited by 46
  22. 2022ERNIE-Search: Bridging Cross-Encoder with Dual-Encoder via Self On-the-fly Distillation for Dense Passage Retrievalcited by 41
  23. 2020Is Retriever Merely an Approximator of Reader?cited by 30
  24. 2022LaPraDoR: Unsupervised Pretrained Dense Retriever for Zero-Shot Text Retrievalcited by 24
  25. 2021Embracing Ambiguity: Shifting the Training Target of NLI Modelscited by 22
  26. 2022DuReader_retrieval: A Large-scale Chinese Benchmark for Passage Retrieval from Web Search Enginecited by 15
  27. 2021Sentence-aware Contrastive Learning for Open-Domain Passage Retrievalcited by 11

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 4; 2023: 4; 2024: 4; 2025: 8; 2026: 1.

21 papers.

Sort:

By area:

  1. 2023Investigating the Factual Knowledge Boundary of Large Language Models with Retrieval Augmentationcited by 146
  2. 2022UniKGQA: Unified Retrieval and Reasoning for Solving Multi-hop Question Answering Over Knowledge Graphcited by 100
  3. 2023Llama2Vec: Unsupervised Adaptation of Large Language Models for Dense Retrievalcited by 77
  4. 2025A Survey of Multimodal Retrieval-Augmented Generationcited by 37
  5. 2024REAR: A Relevance-Aware Retrieval-Augmented Framework for Open-Domain Question Answeringcited by 28
  6. 2024ChatRetriever: Adapting Large Language Models for Generalized and Robust Conversational Dense Retrievalcited by 18
  7. 2023RetroMAE-2: Duplex Masked Auto-Encoder For Pre-Training Retrieval-Oriented Language Modelscited by 13
  8. 2022MASTER: Multi-task Pre-trained Bottlenecked Masked Autoencoders are Better Dense Retrieverscited by 7
  9. 2025ConFit v2: Improving Resume-Job Matching using Hypothetical Resume Embedding and Runner-Up Hard-Negative Miningcited by 4
  10. 2024UniRetriever: Multi-task Candidates Selection for Various Context-Adaptive Conversational Retrievalcited by 3
  11. 2022RetroMAE v2: Duplex Masked Auto-Encoder For Pre-Training Retrieval-Oriented Language Modelscited by 2
  12. 2025MMGeoLM: Hard Negative Contrastive Learning for Fine-Grained Geometric Understanding in Large Multimodal Modelscited by 2
9 more
  1. 2025CoDiEmb: A Collaborative yet Distinct Framework for Unified Representation Learning in Information Retrieval and Semantic Textual Similaritycited by 2
  2. 2022LEAD: Liberal Feature-based Distillation for Dense Retrievalcited by 1
  3. 2023MVMR: A New Framework for Evaluating Faithfulness of Video Moment Retrieval against Multiple Distractorscited by 1
  4. 2024Query-oriented Data Augmentation for Session Searchcited by 0
  5. 2025Intended Target Identification for Anomia Patients with Gradient-based Selective Augmentationcited by 0
  6. 2025RRRA: Resampling and Reranking through a Retriever Adaptercited by 0
  7. 2025GeoGPT-RAG Technical Reportcited by 0
  8. 2025CroPS: Improving Dense Retrieval with Cross-Perspective Positive Samples in Short-Video Searchcited by 0
  9. 2026When Hard Negatives Hurt: Bridging the Generative-Discriminative Gap in Hard Negative Synthesis for Retrievalcited by 0