mlsanityData to 7 Oct 2026

Dense X Retrieval: What Retrieval Granularity Should We Use?

arXiv 2312.06648 ↗First posted 11 Dec 2023cs.CL cs.AI cs.IRcited by 88

Not on the map. Why?

Cited by 48 in Retrieval-augmented generation, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Dense retrieval has become a prominent method to obtain relevant context or world knowledge in open-domain NLP tasks. When we use a learned dense retriever on a retrieval corpus at inference time, an often-overlooked design choice is the retrieval unit in which the corpus is indexed, e.g. document, passage, or sentence. We discover that the retrieval unit choice significantly impacts the performance of both retrieval and downstream tasks. Distinct from the typical approach of using passages or sentences, we introduce a novel retrieval unit, proposition, for dense retrieval. Propositions are defined as atomic expressions within text, each encapsulating a distinct factoid and presented in a concise, self-contained natural language format. We conduct an empirical comparison of different retrieval granularity. Our experiments reveal that indexing a corpus by fine-grained units such as propositions significantly outperforms passage-level units in retrieval tasks. Moreover, constructing prompts with fine-grained retrieved units for retrieval-augmented language models improves the performance of downstream QA tasks given a specific computation budget.

Key facts. Cited by 88 papers in the map (data through 7 Oct 2026). Builds on 51 of its 76 references that are in the map. First posted on arXiv 11 Dec 2023.

Builds on

51 of 76 references.

References · 51

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  3. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
  4. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  5. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  6. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,049
  7. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
  8. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,115
  9. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,779
  10. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  11. 2022Scaling Instruction-Finetuned Language Modelscited by 2,997
  12. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,895
38 more
  1. 2021SimCSE: Simple Contrastive Learning of Sentence Embeddingscited by 2,168
  2. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,800
  3. 2016MS MARCO: A Human Generated MAchine Reading COmprehension Datasetcited by 1,591
  4. 2023FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generationcited by 1,259
  5. 2021BEIR: A Heterogenous Benchmark for Zero-shot Evaluation of Information Retrieval Modelscited by 1,225
  6. 2020Leveraging Passage Retrieval with Generative Models for Open Domain Question Answeringcited by 1,203
  7. 2021Unsupervised Dense Information Retrieval with Contrastive Learningcited by 1,203
  8. 2020ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERTcited by 1,160
  9. 2020Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrievalcited by 846
  10. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 820
  11. 2020KILT: a Benchmark for Knowledge Intensive Language Taskscited by 571
  12. 2021ColBERTv2: Effective and Efficient Retrieval via Lightweight Late Interactioncited by 464
  13. 2021Large Dual Encoders Are Generalizable Retrieverscited by 454
  14. 2020AmbigQA: Answering Ambiguous Open-domain Questionscited by 385
  15. 2022Generate rather than Retrieve: Large Language Models are Strong Context Generatorscited by 296
  16. 2023RECOMP: Improving Retrieval-Augmented LMs with Compression and Selective Augmentationcited by 253
  17. 2021Efficiently Teaching an Effective Dense Retriever with Balanced Topic Aware Samplingcited by 238
  18. 2022Promptagator: Few-shot Dense Retrieval From 8 Examplescited by 233
  19. 2020Sparse, Dense, and Attentional Representations for Text Retrievalcited by 213
  20. 2021Unsupervised Corpus Aware Language Model Pre-training for Dense Passage Retrievalcited by 203
  21. 2019Multilingual Universal Sentence Encoder for Semantic Retrievalcited by 192
  22. 2023Chain-of-Note: Enhancing Robustness in Retrieval-Augmented Language Modelscited by 180
  23. 2020Pre-training Tasks for Embedding-based Large-scale Retrievalcited by 153
  24. 2021Simple Entity-Centric Questions Challenge Dense Retrieverscited by 140
  25. 2021GPL: Generative Pseudo Labeling for Unsupervised Domain Adaptation of Dense Retrievalcited by 119
  26. 2023WiCE: Real-World Entailment for Claims in Wikipediacited by 118
  27. 2023How to Train Your DRAGON: Diverse Augmentation Towards Generalizable Dense Retrievalcited by 116
  28. 2021Decontextualization: Making Sentences Stand-Alonecited by 101
  29. 2021End-to-End Training of Neural Retrievers for Open-Domain Question Answeringcited by 75
  30. 2019Real-Time Open-Domain Question Answering with Dense-Sparse Phrase Indexcited by 64
  31. 2021Salient Phrase Aware Dense Retrieval: Can a Dense Retriever Imitate a Sparse One?cited by 59
  32. 2022UnifiedQA-v2: Stronger Generalization via Broader Cross-Format Trainingcited by 57
  33. 2021Domain-matched Pre-training Tasks for Dense Retrievalcited by 53
  34. 2022Multi-View Document Representation Learning for Open-Domain Dense Retrievalcited by 41
  35. 2021Phrase Retrieval Learns Passage Retrieval, Toocited by 33
  36. 2022PropSegmEnt: A Large-Scale Corpus for Proposition-Level Segmentation and Entailment Recognitioncited by 29
  37. 2021Zero-Shot Dense Retrieval with Momentum Adversarial Domain Invariant Representationscited by 27
  38. 2023Chain-of-Skills: A Configurable Model for Open-domain Question Answeringcited by 19

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 1; 2024: 18; 2025: 30; 2026: 39.

88 papers.

Sort:

By area:

  1. 2023Retrieval-Augmented Generation for Large Language Models: A Surveycited by 2,185
  2. 2024LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memorycited by 656
  3. 2024HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Modelscited by 273
  4. 2024Medical Graph RAG: Towards Safe Medical Large Language Model via Graph Retrieval-Augmented Generationcited by 89
  5. 2024Speculative RAG: Enhancing Retrieval Augmented Generation through Draftingcited by 76
  6. 2024Reliable, Adaptable, and Attributable Language Models with Retrievalcited by 73
  7. 2024Late Chunking: Contextual Chunk Embeddings Using Long-Context Embedding Modelscited by 70
  8. 2024KAG: Boosting LLMs in Professional Domains via Knowledge Augmented Generationcited by 67
  9. 2024Retrieval Augmented Generation (RAG) and Beyond: A Comprehensive Survey on How to Make your LLMs use External Data More Wiselycited by 65
  10. 2025GFM-RAG: Graph Foundation Model for Retrieval Augmented Generationcited by 40
  11. 2024MemLLM: Finetuning LLMs to Use An Explicit Read-Write Memorycited by 35
  12. 2025A Survey on Knowledge-Oriented Retrieval-Augmented Generationcited by 34
38 more
  1. 2024Model Swarms: Collaborative Search to Adapt LLM Experts via Swarm Intelligencecited by 27
  2. 2024SiReRAG: Indexing Similar and Related Information for Multihop Reasoningcited by 23
  3. 2025PropRAG: Guiding Retrieval with Beam Search over Proposition Pathscited by 22
  4. 2025U-NIAH: Unified RAG and LLM Evaluation for Long Context Needle-In-A-Haystackcited by 20
  5. 2025PIKE-RAG: sPecIalized KnowledgE and Rationale Augmented Generationcited by 18
  6. 2025UniversalRAG: Retrieval-Augmented Generation over Corpora of Diverse Modalities and Granularitiescited by 18
  7. 2024Meta-Chunking: Learning Text Segmentation and Semantic Completion via Logical Perceptioncited by 17
  8. 2025Top Ten Challenges Towards Agentic Neural Graph Databasescited by 13
  9. 2024Beyond Relevance: Evaluate and Improve Retrievers on Perspective Awarenesscited by 12
  10. 2025When LLMs Meet API Documentation: Can Retrieval Augmentation Aid Code Generation Just as It Helps Developers?cited by 12
  11. 2025Who Taught the Lie? Responsibility Attribution for Poisoned Knowledge in Retrieval-Augmented Generationcited by 12
  12. 2024Cognitive Kernel: An Open-source Agent System towards Generalist Autopilotscited by 9
  13. 2025A Survey on Retrieval And Structuring Augmented Generation with Large Language Modelscited by 8
  14. 2024Reward-RAG: Enhancing RAG with Reward Driven Supervisioncited by 7
  15. 2025Hierarchical Lexical Graph for Enhanced Multi-Hop Retrievalcited by 7
  16. 2025AGRAG: Advanced Graph-based Retrieval-Augmented Generation for LLMscited by 7
  17. 2024Thread: A Logic-Based Data Organization Paradigm for How-To Question Answering with Retrieval Augmented Generationcited by 6
  18. 2025GORAG: Graph-based Online Retrieval Augmented Generation for Dynamic Few-shot Social Media Text Classificationcited by 6
  19. 2025Passage Segmentation of Documents for Extractive Question Answeringcited by 6
  20. 2025Incorporating Legal Structure in Retrieval-Augmented Generation: A Case Study on Copyright Fair Usecited by 6
  21. 2025PG-Agent: An Agent Powered by Page Graphcited by 6
  22. 2026MemCollab: Cross-Model Memory Collaboration via Contrastive Trajectory Distillationcited by 6
  23. 2024BRIEF: Bridging Retrieval and Inference for Multi-hop Reasoning via Compressioncited by 5
  24. 2025Demo: Healthcare Agent Orchestrator (HAO) for Patient Summarization in Molecular Tumor Boardscited by 5
  25. 2025A Survey of LLM × DATAcited by 4
  26. 2025Scaling External Knowledge Input Beyond Context Windows of LLMs via Multi-Agent Collaborationcited by 4
  27. 2026Taxonomy of the Retrieval System Framework: Pitfalls and Paradigmscited by 4
  28. 2026Experience Transfer for Multimodal LLM Agents in Minecraft Gamecited by 4
  29. 2025Practical Code RAG at Scale: Task-Aware Retrieval Design Choices under Compute Budgetscited by 3
  30. 2026NeocorRAG: Less Irrelevant Information, More Explicit Evidence, and More Effective Recall via Evidence Chainscited by 3
  31. 2026Sem-Detect: Semantic Level Detection of AI Generated Peer-Reviewscited by 3
  32. 2026Search, Inspect, Fetch: Exploiting Structure-Aware Boolean Retrieval for Deep-Search Agentscited by 3
  33. 2025Cooperative Retrieval-Augmented Generation for Question Answering: Mutual Information Exchange and Ranking by Contrasting Layerscited by 2
  34. 2026Retrieving Minimal and Sufficient Reasoning Subgraphs with Graph Foundation Models for Path-aware GraphRAGcited by 2
  35. 2026RL-Index: Reinforcement Learning for Retrieval Index Reasoningcited by 2
  36. 2025Estimating Optimal Context Length for Hybrid Retrieval-augmented Multi-document Summarizationcited by 1
  37. 2025DySK-Attn: A Framework for Efficient, Real-Time Knowledge Updating in Large Language Models via Dynamic Sparse Knowledge Attentioncited by 1
  38. 2025MoM: Mixtures of Scenario-Aware Document Memories for Retrieval-Augmented Generation Systemscited by 1

50 most cited of 88 All 88 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024FunnelRAG: A Coarse-to-Fine Progressive Retrieval Paradigm for RAG
  2. 2024Improve Dense Passage Retrieval with Entailment Tuning
  3. 2026Xetrieval: Mechanistically Explaining Dense Retrieval
  4. 2024Dense Passage Retrieval: Is it Retrieving?
  5. 2025GRITHopper: Decomposition-Free Multi-Hop Dense Retrieval
  6. 2026BM25 Wins at Scale: A Scaling Study of Retrieval-Augmented Generation Paradigms
  7. 2026Uncertainty-Aware Hybrid Retrieval for Long-Document RAG
  8. 2024CoFE-RAG: A Comprehensive Full-chain Evaluation Framework for Retrieval-Augmented Generation with Enhanced Data Diversity