mlsanityData to 14 Oct 2026

MBA-RAG: a Bandit Approach for Adaptive Retrieval-Augmented Generation through Question Complexity

arXiv 2412.01572 ↗First posted 2 Dec 2024cs.AIcited by 29

On the map: Retrieval-augmented generation → Agentic RAG, agents and memory

Abstract (arXiv)

Retrieval Augmented Generation (RAG) has proven to be highly effective in boosting the generative performance of language model in knowledge-intensive tasks. However, existing RAG framework either indiscriminately perform retrieval or rely on rigid single-class classifiers to select retrieval methods, leading to inefficiencies and suboptimal performance across queries of varying complexity. To address these challenges, we propose a reinforcement learning-based framework that dynamically selects the most suitable retrieval strategy based on query complexity. % our solution Our approach leverages a multi-armed bandit algorithm, which treats each retrieval method as a distinct ``arm'' and adapts the selection process by balancing exploration and exploitation. Additionally, we introduce a dynamic reward function that balances accuracy and efficiency, penalizing methods that require more retrieval steps, even if they lead to a correct result. Our method achieves new state of the art results on multiple single-hop and multi-hop datasets while reducing retrieval costs. Our code are available at https://github.com/FUTUREEEEEE/MBA .

Key facts. Cited by 29 papers in the map (data through 14 Oct 2026). Placed in Agentic RAG, agents and memory, one of 15 directions in Retrieval-augmented generation. Builds on 13 of its 31 references that are in the map. First posted on arXiv 2 Dec 2024.

Builds on

13 of 31 references.

References · 13

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  2. 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,307
  3. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,148
  4. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  5. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  6. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,955
  7. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,839
  8. 2020Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Stepscited by 1,282
  9. 2022When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memoriescited by 1,128
  10. 2022Interleaving Retrieval with Chain-of-Thought Reasoning for Knowledge-Intensive Multi-Step Questionscited by 923
  11. 2024Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexitycited by 451
  12. 2024CO2: Efficient Distributed Training with Full Communication-Computation Overlapcited by 17
1 more
  1. 2023Spatial-Temporal Interplay in Human Mobility: A Hierarchical Reinforcement Learning Approach with Hypergraph Representationcited by 6

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 12; 2026: 16.

29 papers.

Sort:

By area:

  1. 2025A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Modelscited by 88
  2. 2025When to use Graphs in RAG: A Comprehensive Analysis for Graph Retrieval-Augmented Generationcited by 48
  3. 2025KABB: Knowledge-Aware Bayesian Bandits for Dynamic Expert Coordination in Multi-Agent Systemscited by 37
  4. 2025UniversalRAG: Retrieval-Augmented Generation over Corpora of Diverse Modalities and Granularitiescited by 20
  5. 2024METIS: Fast Quality-Aware RAG Systems with Configuration Adaptationcited by 17
  6. 2025LAG: Logic-Augmented Generation from a Cartesian Perspectivecited by 10
  7. 2025Fast or Better? Balancing Accuracy and Cost in Retrieval-Augmented Generation with Flexible User Controlcited by 6
  8. 2026RAGRouter-Bench: A Dataset and Benchmark for Adaptive RAG Routingcited by 6
  9. 2026Bayesian Active Learning with Gaussian Processes Guided by LLM Relevance Scoring for Dense Passage Retrievalcited by 6
  10. 2026BalanceRAG: Joint Risk Calibration for Cascaded Retrieval-Augmented Generationcited by 4
  11. 2025OkraLong: A Flexible Retrieval-Augmented Framework for Long-Text Query Processingcited by 2
  12. 2025KAG-Thinker: Interactive Thinking and Deep Reasoning in LLMs via Knowledge-Augmented Generationcited by 2
17 more
  1. 2025Confident RAG: Enhancing the Performance of LLMs for Mathematics Question Answering through Multi-Embedding and Confidence Scoringcited by 2
  2. 2025Online-Optimized RAG for Tool Use and Function Callingcited by 2
  3. 2026RAGTrack: Language-aware RGBT Tracking with Retrieval-Augmented Generationcited by 2
  4. 2026Budget-Constrained Online Retrieval-Augmented Generation: The Chunk-as-a-Service Modelcited by 2
  5. 2026Maximizing Rollout Informativeness under a Fixed Budget: A Submodular View of Tree Search for Tool-Use Agentic Reinforcement Learningcited by 2
  6. 2025POLYRAG: Integrating Polyviews into Retrieval-Augmented Generation for Medical Applicationscited by 1
  7. 2025Feedback-Induced Performance Decline in LLM-Based Decision-Makingcited by 1
  8. 2026A Component-Based Survey of Interactions between Large Language Models and Multi-Armed Banditscited by 1
  9. 2026Diagnosing and Repairing Factual Errors in RAG under Budget Constraintscited by 1
  10. 2026A Dynamic Retrieval-Augmented Generation System with Selective Memory and Remembrancecited by 0
  11. 2026Deep GraphRAG: A Balanced Approach to Hierarchical Retrieval and Adaptive Integrationcited by 0
  12. 2026Is a Picture Worth a Thousand Words? Adaptive Multimodal Fact-Checking with Visual Evidence Necessitycited by 0
  13. 2026MAB-DQA: Addressing Query Aspect Importance in Document Question Answering with Multi-Armed Banditscited by 0
  14. 2026ADEPT: An Entropy-Driven Dual-Strategy Agent for Interactive Video Retrievalcited by 0
  15. 2026LODESTAR: Robust Entropy-Based Answer Selection in Retrieval-Augmented Generation for Question Answering -- Directing Frozen-LLM Entropy with a Reinforcement-Learned Prompt Polarizer under Misleading Passagescited by 0
  16. 2026Beyond the Query: Do Retrieval Signals Improve Adaptive Multimodal RAG Routing?cited by 0
  17. 2026Fusion is the New Mutation: Bandit-Guided Evolution on Workflow Graphscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Adapting to Non-Stationary Environments: Multi-Armed Bandit Enhanced Retrieval-Augmented Generation on Knowledge Graphs
  2. 2025Curriculum Guided Reinforcement Learning for Efficient Multi Hop Retrieval Augmented Generation
  3. 2025Query Decomposition for RAG: Balancing Exploration-Exploitation
  4. 2026Understanding the Behaviors of Environment-aware Information Retrieval
  5. 2025Multi-hop Reasoning via Early Knowledge Alignment
  6. 2025Towards Adaptive Memory-Based Optimization for Enhanced Retrieval-Augmented Generation
  7. 2025Stop-RAG: Value-Based Retrieval Control for Iterative RAG
  8. 2026One Size Does Not Fit All! Dynamic Retriever and Generator Selection for RAG