mlsanityData to 14 Oct 2026

Improving Query Representations for Dense Retrieval with Pseudo Relevance Feedback

arXiv 2108.13454 ↗First posted 30 Aug 2021cs.IR cs.AIcited by 47

Not on the map. Why?

Abstract (arXiv)

Dense retrieval systems conduct first-stage retrieval using embedded representations and simple similarity metrics to match a query to documents. Its effectiveness depends on encoded embeddings to capture the semantics of queries and documents, a challenging task due to the shortness and ambiguity of search queries. This paper proposes ANCE-PRF, a new query encoder that uses pseudo relevance feedback (PRF) to improve query representations for dense retrieval. ANCE-PRF uses a BERT encoder that consumes the query and the top retrieved documents from a dense retrieval model, ANCE, and it learns to produce better query embeddings directly from relevance labels. It also keeps the document index unchanged to reduce overhead. ANCE-PRF significantly outperforms ANCE and other recent dense retrieval systems on several datasets. Analysis shows that the PRF encoder effectively captures the relevant and complementary information from PRF documents, while ignoring the noise with its learned attention mechanism.

Key facts. Cited by 47 papers in the map (data through 14 Oct 2026). Builds on 25 of its 36 references that are in the map. First posted on arXiv 30 Aug 2021.

Builds on

25 of 36 references.

References · 25

  1. 2017Attention Is All You Needcited by 60,361
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  3. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  4. 2017Billion-scale similarity search with GPUscited by 2,497
  5. 2016MS MARCO: A Human Generated MAchine Reading COmprehension Datasetcited by 1,605
  6. 2021BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Modelscited by 1,252
  7. 2020ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERTcited by 1,178
  8. 2020Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrievalcited by 849
  9. 2019Latent Retrieval for Weakly Supervised Open Domain Question Answeringcited by 695
  10. 2020Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillationcited by 593
  11. 2020Overview of the TREC 2019 deep learning trackcited by 286
  12. 2021Efficiently Teaching an Effective Dense Retriever with Balanced Topic Aware Samplingcited by 238
13 more
  1. 2020Sparse, Dense, and Attentional Representations for Text Retrievalcited by 214
  2. 2021Overview of the TREC 2020 deep learning trackcited by 196
  3. 2017A Deep Relevance Matching Model for Ad-hoc Retrievalcited by 126
  4. 2019Learning Dense Representations for Entity Retrievalcited by 122
  5. 2017End-to-End Neural Ad-hoc Ranking with Kernel Poolingcited by 85
  6. 2018Learning a Deep Listwise Context Model for Ranking Refinementcited by 62
  7. 2020Distilling Dense Representations for Ranking using Tightly-Coupled Teacherscited by 52
  8. 2021Pseudo-Relevance Feedback for Multiple Representation Dense Retrievalcited by 32
  9. 2021How Deep is your Learning: the DL-HARD Annotated Deep Learning Datasetcited by 18
  10. 2018NPRF: A Neural Pseudo Relevance Feedback Framework for Ad-hoc Information Retrievalcited by 17
  11. 2020Learning To Retrieve: How to Train a Dense Retrieval Model Effectively and Efficientlycited by 11
  12. 2021Co-BERT: A Context-Aware BERT Retrieval Model Incorporating Local and Query-specific Contextcited by 6
  13. 2021PGT: Pseudo Relevance Feedback Using a Graph-Based Transformercited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 6; 2023: 8; 2024: 10; 2025: 11; 2026: 12.

47 papers.

Sort:

By area:

  1. 2023TCRA-LLM: Token Compression Retrieval Augmented Large Language Model for Inference Cost Reductioncited by 34
  2. 2022RACE: Retrieval-Augmented Commit Message Generationcited by 27
  3. 2023ReGen: Zero-Shot Text Classification via Training Data Generation with Progressive Dense Retrievalcited by 26
  4. 2024GenQREnsemble: Zero-Shot LLM Ensemble Prompting for Generative Query Reformulationcited by 23
  5. 2024Utilizing BERT for Information Retrieval: Survey, Applications, Resources, and Challengescited by 20
  6. 2023Expand, Rerank, and Retrieve: Query Reranking for Open-Domain Question Answeringcited by 15
  7. 2024Taxonomy-guided Semantic Indexing for Academic Paper Searchcited by 12
  8. 2025Guided Query Refinement: Multimodal Hybrid Retrieval with Test-Time Optimizationcited by 12
  9. 2024Improving Retrieval in Theme-specific Applications using a Corpus Topical Taxonomycited by 11
  10. 2025Scientific Paper Retrieval with LLM-Guided Semantic-Based Rankingcited by 10
  11. 2022Neural Rankers for Effective Screening Prioritisation in Medical Systematic Review Literature Searchcited by 9
  12. 2022Optimizing Test-Time Query Representations for Dense Retrievalcited by 7
35 more
  1. 2024Generalized Contrastive Learning for Multi-Modal Retrieval and Rankingcited by 7
  2. 2025Improving Scientific Document Retrieval with Concept Coverage-based Query Set Generationcited by 7
  3. 2023Augmenting Zero-Shot Dense Retrievers with Plug-in Mixture-of-Memoriescited by 6
  4. 2024DREQ: Document Re-Ranking Using Entity-based Query Understandingcited by 5
  5. 2024Generative Query Reformulation Using Ensemble Prompting, Document Fusion, and Relevance Feedbackcited by 5
  6. 2022Incorporating Relevance Feedback for Information-Seeking Retrieval using Few-Shot Document Re-Rankingcited by 4
  7. 2023Relevance Feedback with Brain Signalscited by 4
  8. 2024QAEA-DR: A Unified Text Augmentation Framework for Dense Retrievalcited by 4
  9. 2025Pseudo Relevance Feedback is Enough to Close the Gap Between Small and Large Dense Retrieval Modelscited by 4
  10. 2025E2Rank: Unifying Text Embedding and Listwise Reranking for Effective and Efficient Searchcited by 4
  11. 2025Test-time Corpus Feedback: From Retrieval to RAGcited by 3
  12. 2026Beyond Single-Shot: Multi-step Tool Retrieval via Query Planningcited by 3
  13. 2022Asyncval: A Toolkit for Asynchronously Validating Dense Retriever Checkpoints during Trainingcited by 2
  14. 2023ReFIT: Relevance Feedback from a Reranker during Inferencecited by 2
  15. 2025LLM-VPRF: Large Language Model Based Vector Pseudo Relevance Feedbackcited by 2
  16. 2026Succeeding at Scale: Enterprise Retrieval Benchmark Construction and Index-Preserving Query Adaptation for Multi-Tenant Searchcited by 2
  17. 2022LoL: A Comparative Regularization Loss over Query Reformulation Losses for Pseudo-Relevance Feedbackcited by 1
  18. 2023Cross-Model Comparative Loss for Enhancing Neuronal Utility in Language Understandingcited by 1
  19. 2024Adaptive Retrieval and Scalable Indexing for k-NN Search with Cross-Encoderscited by 1
  20. 2026Document Optimization for Black-Box Retrieval via Reinforcement Learningcited by 1
  21. 2026Reproduction Beyond Benchmarks: ConstBERT and ColBERT-v2 Across Backends and Query Distributionscited by 1
  22. 2023Online Distillation for Pseudo-Relevance Feedbackcited by 0
  23. 2024Selecting Query-bag as Pseudo Relevance Feedback for Information-seeking Conversationscited by 0
  24. 2025Multi-Modality Transformer for E-Commerce: Inferring User Purchase Intention to Bridge the Query-Product Gapcited by 0
  25. 2025Intended Target Identification for Anomia Patients with Gradient-based Selective Augmentationcited by 0
  26. 2025CLAP: Coreference-Linked Augmentation for Passage Retrievalcited by 0
  27. 2025REGENT: Relevance-Guided Attention for Entity-Aware Multi-Vector Neural Re-Rankingcited by 0
  28. 2026Improving Scientific Document Retrieval with Academic Concept Indexcited by 0
  29. 2026Task-Adaptive Embedding Refinement via Test-time LLM Guidancecited by 0
  30. 2026Generative Retrieval via Diffusion Transformer with Metric-Ordered Sequence Training and Hybrid-Policy Preference Optimizationcited by 0
  31. 2026PLAID-PRF: Pseudo-Relevance Feedback with Centroid-like Tokens in PLAIDcited by 0
  32. 2026EviReform: Evidence-Guided Query Reformulation for Multi-Hop Graph Retrievalcited by 0
  33. 2026Autoregressive Retriever: Improving Query Understanding from Item Feedback for Universal Multimodal Retrievalcited by 0
  34. 2026Graph-free Retrieval with Graph-Modified Embeddingscited by 0
  35. 2026Cascaded Retrieval with Cross-Modal Query Refinement for Efficient Visual Document Searchcited by 0