mlsanityData to 13 Oct 2026

Chain-of-Retrieval Augmented Generation

arXiv 2501.14342 ↗First posted 24 Jan 2025cs.IR cs.CLcited by 49

On the map: LLM reasoning → Knowledge- and tool-augmented reasoning

also in 1 more
Abstract (arXiv)

This paper introduces an approach for training o1-like RAG models that retrieve and reason over relevant information step by step before generating the final answer. Conventional RAG methods usually perform a single retrieval step before the generation process, which limits their effectiveness in addressing complex queries due to imperfect retrieval results. In contrast, our proposed method, CoRAG (Chain-of-Retrieval Augmented Generation), allows the model to dynamically reformulate the query based on the evolving state. To train CoRAG effectively, we utilize rejection sampling to automatically generate intermediate retrieval chains, thereby augmenting existing RAG datasets that only provide the correct final answer. At test time, we propose various decoding strategies to scale the model's test-time compute by controlling the length and number of sampled retrieval chains. Experimental results across multiple benchmarks validate the efficacy of CoRAG, particularly in multi-hop question answering tasks, where we observe more than 10 points improvement in EM score compared to strong baselines. On the KILT benchmark, CoRAG establishes a new state-of-the-art performance across a diverse range of knowledge-intensive tasks. Furthermore, we offer comprehensive analyses to understand the scaling behavior of CoRAG, laying the groundwork for future research aimed at developing factual and grounded foundation models.

Key facts. Cited by 49 papers in the map (data through 13 Oct 2026). Placed in Knowledge- and tool-augmented reasoning, one of 17 directions in LLM reasoning. Builds on 39 of its 46 references that are in the map. First posted on arXiv 24 Jan 2025.

Builds on

39 of 46 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025Search-o1: Agentic Search-Enhanced Large Reasoning Models9 Jan
  2. 2023Retrieval-Augmented Generation for Large Language Models: A Survey18 Dec
  3. 2022Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questions20 Dec

Other references · 36

  1. 2024The Llama 3 Herd of Modelscited by 14,486
  2. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,376
  3. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,221
  4. 2024Qwen2.5 Technical Reportcited by 4,840
  5. 2024GPT-4o System Cardcited by 4,536
  6. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,133
  7. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,797
  8. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,716
  9. 2024Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextcited by 3,197
  10. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,181
  11. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,925
  12. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,815
24 more
  1. 2021WebGPT: Browser-assisted question-answering with human feedbackcited by 1,736
  2. 2020Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Stepscited by 1,266
  3. 2022Text Embeddings by Weakly-Supervised Contrastive Pre-trainingcited by 1,229
  4. 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,063
  5. 2022Atlas: Few-shot Learning with Retrieval Augmented Language Modelscited by 947
  6. 2022STaR: Bootstrapping Reasoning With Reasoningcited by 913
  7. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 902
  8. 2020KILT: a Benchmark for Knowledge Intensive Language Taskscited by 571
  9. 2023Active Retrieval Augmented Generationcited by 510
  10. 2023Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervisioncited by 454
  11. 2023Enhancing Retrieval-Augmented Large Language Models with Iterative Retrieval-Generation Synergycited by 344
  12. 2025rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinkingcited by 309
  13. 2023Improving Text Embeddings with Large Language Modelscited by 297
  14. 2023Fine-Tuning LLaMA for Multi-Stage Text Retrievalcited by 271
  15. 2023RA-DIT: Retrieval-Augmented Dual Instruction Tuningcited by 195
  16. 2024FlashRAG: A Modular Toolkit for Efficient Retrieval-Augmented Generation Researchcited by 165
  17. 2022Autoregressive Search Engines: Generating Substrings as Document Identifierscited by 112
  18. 2024LongRAG: Enhancing Retrieval-Augmented Generation with Long-context LLMscited by 106
  19. 2024From Matching to Generation: A Survey on Generative Information Retrievalcited by 89
  20. 2024Inference Scaling for Long-Context Retrieval Augmented Generationcited by 65
  21. 2017Ask the Right Questions: Active Question Reformulation with Reinforcement Learningcited by 44
  22. 2021Multi-task Retrieval for Knowledge-Intensive Taskscited by 44
  23. 2024Auto-RAG: Autonomous Retrieval-Augmented Generation for Large Language Modelscited by 42
  24. 2022Multi-Task Retrieval-Augmented Text Generation with Relevance Samplingcited by 10

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2025: 25; 2026: 24.

49 papers.

Sort:

By area:

  1. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 383
  2. 2025R1-Searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learningcited by 329
  3. 2025WebThinker: Empowering Large Reasoning Models with Deep Research Capabilitycited by 322
  4. 2025Deep Research Agents: A Systematic Examination And Roadmapcited by 126
  5. 2025StepSearch: Igniting LLMs Search Ability via Step-Wise Proximal Policy Optimizationcited by 85
  6. 2025ReasonIR: Training Retrievers for Reasoning Taskscited by 77
  7. 2025SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesiscited by 55
  8. 2025Synergizing RAG and Reasoning: A Systematic Reviewcited by 27
  9. 2025Reasoning Beyond Limits: Advances and Open Problems for LLMscited by 16
  10. 2025Generative AI Act II: Test Time Scaling Drives Cognition Engineeringcited by 14
  11. 2025Stop-RAG: Value-Based Retrieval Control for Iterative RAGcited by 14
  12. 2026Reasoning in Trees: Improving Retrieval-Augmented Generation for Multi-Hop Question Answeringcited by 13
37 more
  1. 2025RARE: Retrieval-Augmented Reasoning Modelingcited by 11
  2. 2026Coding Agents are Effective Long-Context Processorscited by 11
  3. 2025AceSearcher: Bootstrapping Reasoning and Search for LLMs via Reinforced Self-Playcited by 10
  4. 2025HiPRAG: Hierarchical Process Rewards for Efficient Agentic Retrieval Augmented Generationcited by 10
  5. 2025Beneficial Reasoning Behaviors in Agentic Search and Effective Post-training to Obtain Themcited by 7
  6. 2026Mnemis: Dual-Route Retrieval on Hierarchical Graphs for Long-Term LLM Memorycited by 7
  7. 2026How to Train Your Deep Research Agent? Prompt, Reward, and Policy Optimization in Search-R1cited by 6
  8. 2025Scaling Test-Time Inference with Policy-Optimized, Dynamic Retrieval-Augmented Generation via KV Caching and Decodingcited by 5
  9. 2025MetaAgent: Toward Self-Evolving Agent via Tool Meta-Learningcited by 5
  10. 2026Fine-Tuning vs. RAG for Multi-Hop Question Answering with Novel Knowledgecited by 4
  11. 2026URAG: A Benchmark for Uncertainty Quantification in Retrieval-Augmented Large Language Modelscited by 3
  12. 2026Experience as a Compass: Multi-agent RAG with Evolving Orchestration and Agent Promptscited by 3
  13. 2025FrugalRAG: Less is More in RL Finetuning for Multi-Hop Question Answeringcited by 2
  14. 2026Reinforcement Fine-Tuning for History-Aware Dense Retriever in RAGcited by 2
  15. 2026AgenticRag-R1: Agentic Reinforcement Learning with Stack Memory for Multi-Step Reasoning, Retrieval and Memorizingcited by 2
  16. 2025TrustRAG: An Information Assistant with Retrieval Augmented Generationcited by 1
  17. 2025RAISE: Enhancing Scientific Reasoning in LLMs via Step-by-Step Retrievalcited by 1
  18. 2025PrismRAG: Boosting RAG Factuality with Distractor Resilience and Strategized Reasoningcited by 1
  19. 2025Teaching LLMs to Think Mathematically: A Critical Study of Decision-Making via Optimizationcited by 1
  20. 2025An Open and Reproducible Deep Research Agent for Long-Form Question Answeringcited by 1
  21. 2026MG2-RAG: Multi-Granularity Graph for Multimodal Retrieval-Augmented Generationcited by 1
  22. 2026SpecHop: Continuous Speculation for Accelerating Multi-Hop Retrieval Agentscited by 1
  23. 2026Track, Rank, Crack: Epistemic Working Memory Scales Multi-Hop Reasoning in Language Agentscited by 1
  24. 2026Deep Agentic Search for Repository-Level Code Question Answering: An Empirical Studycited by 1
  25. 2026Predicting Partial Answer Quality and Utility in Agentic Retrieval-Augmented Generationcited by 1
  26. 2025DTKG: Dual-Track Knowledge Graph-Verified Reasoning Framework for Multi-Hop QAcited by 0
  27. 2025BambooKG: A Neurobiologically-inspired Frequency-Weight Knowledge Graphcited by 0
  28. 2026RAG Strategies for Natural Language-Based SQL Query and REST API Call Generationcited by 0
  29. 2026Chain of Evidence: Pixel-Level Visual Attribution for Iterative Retrieval-Augmented Generationcited by 0
  30. 2026The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modescited by 0
  31. 2026PROGRESS: Coverage-guided RL to Train Search-augmented LLM Agentcited by 0
  32. 2026Search-GRT: Guided Retrieval Training of Search Agents to Optimize for Complex Question Answeringcited by 0
  33. 2026DreamBench-SWE: A Multi-Session Memory-Hygiene Benchmark for Software Agentscited by 0
  34. 2026CacheRepair: Learning to Repair Cross-Chunk Context in RAG for KV Cache Fusioncited by 0
  35. 2026PrimeSeeker: Capability-Oriented Supervision for Deep Search Agentscited by 0
  36. 2026Lost in Conversation or Lost in Translation? Diagnosing Multi-Turn Degradation in RAGcited by 0
  37. 2026Mapping the RAG Landscape: A Four Axis Taxonomy of Efficiency, Defense, Interactivity, and Reasoningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026AB-RAG: Adaptive Budgeted Retrieval-Augmented Generation for Reliable Question Answering
  2. 2024Multi-Meta-RAG: Improving RAG for Multi-Hop Queries using Database Filtering with LLM-Extracted Metadata
  3. 2025Accelerating Adaptive Retrieval Augmented Generation via Instruction-Driven Representation Reduction of Retrieval Overlaps
  4. 2024ARAGOG: Advanced RAG Output Grading
  5. 2024MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries
  6. 2025CoRAG: Collaborative Retrieval-Augmented Generation
  7. 2025DACL-RAG: Data Augmentation Strategy with Curriculum Learning for Retrieval-Augmented Generation
  8. 2025A Comparison of Independent and Joint Fine-tuning Strategies for Retrieval-Augmented Generation