mlsanityData to 7 Oct 2026

Stochastic RAG: End-to-End Retrieval-Augmented Generation through Expected Utility Maximization

arXiv 2405.02816 ↗First posted 5 May 2024cs.CL cs.IR cs.LGcited by 29

On the map: Retrieval-augmented generation → RAG pipeline design and evaluation

Abstract (arXiv)

This paper introduces Stochastic RAG--a novel approach for end-to-end optimization of retrieval-augmented generation (RAG) models that relaxes the simplifying assumptions of marginalization and document independence, made in most prior work. Stochastic RAG casts the retrieval process in RAG as a stochastic sampling without replacement process. Through this formulation, we employ straight-through Gumbel-top-k that provides a differentiable approximation for sampling without replacement and enables effective end-to-end optimization for RAG. We conduct extensive experiments on seven diverse datasets on a wide range of tasks, from open-domain question answering to fact verification to slot-filling for relation extraction and to dialogue systems. By applying this optimization method to a recent and effective RAG model, we advance state-of-the-art results on six out of seven datasets.

Key facts. Cited by 29 papers in the map (data through 7 Oct 2026). Placed in RAG pipeline design and evaluation, one of 15 directions in Retrieval-augmented generation. Builds on 37 of its 51 references that are in the map. First posted on arXiv 5 May 2024.

Builds on

37 of 51 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Evaluating Retrieval Quality in Retrieval-Augmented Generation21 Apr
  2. 2024Optimization Methods for Personalizing Large Language Models through Retrieval Augmentation9 Apr
  3. 2022FiD-Light: Efficient and Effective Retrieval-Augmented Text Generation28 Sep
  4. 2022A Survey on Retrieval-Augmented Text Generation2 Feb

Other references · 33

  1. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  2. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  3. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,779
  4. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,686
  5. 2014Sequence to Sequence Learning with Neural Networkscited by 2,974
  6. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,895
  7. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  8. 2021WebGPT: Browser-assisted question-answering with human feedbackcited by 1,722
  9. 2016MS MARCO: A Human Generated MAchine Reading COmprehension Datasetcited by 1,591
  10. 2018FEVER: a large-scale dataset for Fact Extraction and VERificationcited by 1,362
  11. 2018Adafactor: Adaptive Learning Rates with Sublinear Memory Costcited by 896
  12. 2020Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrievalcited by 846
21 more
  1. 2023REPLUG: Retrieval-Augmented Black-Box Language Modelscited by 786
  2. 2020KILT: a Benchmark for Knowledge Intensive Language Taskscited by 571
  3. 2018Wizard of Wikipedia: Knowledge-Powered Conversational agentscited by 530
  4. 2017Zero-Shot Relation Extraction via Reading Comprehensioncited by 425
  5. 2023LaMP: When Large Language Models Meet Personalizationcited by 425
  6. 2023On the Planning Abilities of Large Language Models : A Critical Investigationcited by 308
  7. 2023FreshLLMs: Refreshing Large Language Models with Search Engine Augmentationcited by 308
  8. 2020Distilling Knowledge from Reader to Retriever for Question Answeringcited by 226
  9. 2019Stochastic Beams and Where to Find Them: The Gumbel-Top-k Trick for Sampling Sequences Without Replacementcited by 205
  10. 2021Retrieving and Reading: A Comprehensive Survey on Open-domain Question Answeringcited by 178
  11. 2022Autoregressive Search Engines: Generating Substrings as Document Identifierscited by 112
  12. 2018Retrieve and Refine: Improved Sequence Generation Models For Dialoguecited by 82
  13. 2021End-to-End Training of Neural Retrievers for Open-Domain Question Answeringcited by 75
  14. 2022Retrieval-Enhanced Machine Learningcited by 46
  15. 2021Hindsight: Posterior-guided training of retrievers for improved open-ended generationcited by 38
  16. 2020Rao-Blackwellizing the Straight-Through Gumbel-Softmax Gradient Estimatorcited by 37
  17. 2021PLATO-XL: Exploring the Large-scale Pre-training of Dialogue Generationcited by 37
  18. 2021Robust Retrieval Augmented Generation for Zero-shot Slot Fillingcited by 24
  19. 2023GripRank: Bridging the Gap between Retrieval and Generation via the Generative Knowledge Improved Passage Rankingcited by 11
  20. 2022Multi-Task Retrieval-Augmented Text Generation with Relevance Samplingcited by 10
  21. 2024Re3val: Reinforced and Reranked Generative Retrievalcited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 9; 2025: 14; 2026: 6.

29 papers.

Sort:

By area:

  1. 2024Searching for Best Practices in Retrieval-Augmented Generationcited by 83
  2. 2025Improving Retrieval-Augmented Generation through Multi-Agent Reinforcement Learningcited by 39
  3. 2025Retrieval-Augmented Generation: A Comprehensive Survey of Architectures, Enhancements, and Robustness Frontierscited by 26
  4. 2024Claim Verification in the Age of Large Language Models: A Surveycited by 23
  5. 2024Towards Fair RAG: On the Impact of Fair Ranking in Retrieval-Augmented Generationcited by 15
  6. 2024Learning to Rank for Multiple Retrieval-Augmented Models through Iterative Utility Maximizationcited by 14
  7. 2024Accelerating Retrieval-Augmented Generationcited by 14
  8. 2025Training a Utility-based Retriever Through Shared Context Attribution for Retrieval-Augmented Language Modelscited by 13
  9. 2024Automatic Database Configuration Debugging using Retrieval-Augmented Language Modelscited by 12
  10. 2024Retrieval-Enhanced Machine Learning: Synthesis and Opportunitiescited by 11
  11. 2025Iterative Self-Incentivization Empowers Large Language Models as Agentic Searcherscited by 11
  12. 2026Agentic-R: Learning to Retrieve for Agentic Searchcited by 10
17 more
  1. 2025Direct Retrieval-augmented Optimization: Synergizing Knowledge Selection and Language Modelscited by 9
  2. 2025Mitigating Bias in RAG: Controlling the Embeddercited by 6
  3. 2025The Next Phase of Scientific Fact-Checking: Advanced Evidence Retrieval from Complex Structured Academic Paperscited by 6
  4. 2025Utility-Focused LLM Annotation for Retrieval and Retrieval-Augmented Generationcited by 5
  5. 2025LLM-Specific Utility for Retrieval-Augmented Generationcited by 4
  6. 2026Learning to Retrieve from Agent Trajectoriescited by 3
  7. 2026How Retrieved Context Shapes Internal Representations in RAGcited by 2
  8. 2024Binary Hypothesis Testing for Softmax Models and Leverage Score Modelscited by 1
  9. 2025Oreo: A Plug-in Context Reconstructor to Enhance Retrieval-Augmented Generationcited by 1
  10. 2025RAGtifier: Evaluating RAG Generation Approaches of State-of-the-Art RAG Systems for the SIGIR LiveRAG Competitioncited by 1
  11. 2025Improving End-to-End Training of Retrieval-Augmented Generation Models via Joint Stochastic Approximationcited by 1
  12. 2026Reinforcement Fine-Tuning for History-Aware Dense Retriever in RAGcited by 1
  13. 2026FLOWREADER: Min-Cost Flow Optimization for Multi-Modal Long Document Q&Acited by 1
  14. 2024Leveraging Retrieval-Augmented Generation for Persian University Knowledge Retrievalcited by 0
  15. 2025Towards Efficient Educational Chatbots: Benchmarking RAG Frameworkscited by 0
  16. 2025Plan-and-Refine: Diverse and Comprehensive Retrieval-Augmented Generationcited by 0
  17. 2026Critic-R: Improving Agentic Search using Instruction-tuned Retrievers with Natural Language Introspective Feedbackcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Toward Optimal Search and Retrieval for RAG
  2. 2024A Comprehensive Survey of Retrieval-Augmented Generation (RAG): Evolution, Current Landscape and Future Directions
  3. 2025A Systematic Review of Key Retrieval-Augmented Generation (RAG) Systems: Progress, Gaps, and Future Directions
  4. 2024Towards Optimizing a Retrieval Augmented Generation using Large Language Model on Academic Data
  5. 2024Evaluation of Retrieval-Augmented Generation: A Survey
  6. 2025Question Decomposition for Retrieval-Augmented Generation
  7. 2025A Survey on Knowledge-Oriented Retrieval-Augmented Generation
  8. 2025mmRAG: A Modular Benchmark for Retrieval-Augmented Generation over Text, Tables, and Knowledge Graphs