mlsanityData to 14 Oct 2026

AttentionRAG: Attention-Guided Context Pruning in Retrieval-Augmented Generation

arXiv 2503.10720 ↗First posted 13 Mar 2025cs.CL cs.AIcited by 19

On the map: Retrieval-augmented generation → Generation and context integration

Abstract (arXiv)

While RAG demonstrates remarkable capabilities in LLM applications, its effectiveness is hindered by the ever-increasing length of retrieved contexts, which introduces information redundancy and substantial computational overhead. Existing context pruning methods, such as LLMLingua, lack contextual awareness and offer limited flexibility in controlling compression rates, often resulting in either insufficient pruning or excessive information loss. In this paper, we propose AttentionRAG, an attention-guided context pruning method for RAG systems. The core idea of AttentionRAG lies in its attention focus mechanism, which reformulates RAG queries into a next-token prediction paradigm. This mechanism isolates the query's semantic focus to a single token, enabling precise and efficient attention calculation between queries and retrieved contexts. Extensive experiments on LongBench and Babilong benchmarks show that AttentionRAG achieves up to 6.3× context compression while outperforming LLMLingua methods by around 10% in key metrics.

Key facts. Cited by 19 papers in the map (data through 14 Oct 2026). Placed in Generation and context integration, one of 15 directions in Retrieval-augmented generation. Builds on 26 of its 36 references that are in the map. First posted on arXiv 13 Mar 2025.

Builds on

26 of 36 references.

Closest ancestors · 5

Same direction, newest first

  1. 2024Contextual Compression in Retrieval-Augmented Generation for Large Language Models: A Survey20 Sep
  2. 2024Accelerating Inference of Retrieval-Augmented Generation via Sparse Context Selection25 May
  3. 2024xRAG: Extreme Context Compression for Retrieval-augmented Generation with One Token22 May
  4. 2023RECOMP: Improving Retrieval-Augmented LMs with Compression and Selective Augmentation6 Oct
  5. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks22 May

Other references · 21

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,148
  3. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  4. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,955
  5. 2023LongBench: A Bilingual, Multitask Benchmark for Long Context Understandingcited by 1,417
  6. 2020Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Stepscited by 1,282
  7. 2024M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillationcited by 1,073
  8. 2019What Does BERT Look At? An Analysis of BERT's Attentioncited by 1,003
  9. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 841
  10. 2023Learning to Compress Prompts with Gist Tokenscited by 390
  11. 2024LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compressioncited by 348
  12. 2021Efficient Attentions for Long Document Summarizationcited by 345
9 more
  1. 2024BABILong: Testing the Limits of LLMs with Long Context Reasoning-in-a-Haystackcited by 215
  2. 2024Leave No Document Behind: Benchmarking Long-Context LLMs with Extended Multi-Doc QAcited by 122
  3. 2022Overcoming a Theoretical Limitation of Self-Attentioncited by 113
  4. 2024Found in the Middle: Calibrating Positional Attention Bias Improves Long Context Utilizationcited by 101
  5. 2025ViDoRAG: Visual Document Retrieval-Augmented Generation via Dynamic Iterative Reasoning Agentscited by 66
  6. 2024RAG-QA Arena: Evaluating Domain Robustness for Long-form Retrieval Augmented Question Answeringcited by 43
  7. 2024Attend First, Consolidate Later: On the Importance of Attention in Different LLM Layerscited by 25
  8. 2024RAG-Instruct: Boosting LLMs with Diverse Retrieval-Augmented Instructionscited by 25
  9. 2024Enhancing Retrieval and Managing Retrieval: A Four-Module Synergy for Improved Quality and Efficiency in RAG Systemscited by 15

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 16.

19 papers.

Sort:

By area:

  1. 2025VRAG-RL: Empower Vision-Perception-Based RAG for Visually Rich Information Understanding via Iterative Reasoning with Reinforcement Learningcited by 68
  2. 2026SWE-Pruner: Self-Adaptive Context Pruning for Coding Agentscited by 44
  3. 2025LongCodeZip: Compress Long Context for Code Language Modelscited by 39
  4. 2026Enhancing Financial Report Question-Answering: A Retrieval-Augmented Generation System with Reranking Analysiscited by 21
  5. 2025RaCoT: Plug-and-Play Contrastive Example Generation Mechanism for Enhanced LLM Reasoning Reliabilitycited by 14
  6. 2026Reasoning in Trees: Improving Retrieval-Augmented Generation for Multi-Hop Question Answeringcited by 14
  7. 2026COMI: Coarse-to-fine Context Compression via Marginal Information Gaincited by 13
  8. 2026Sustainable Hybrid Document-Routed Retrieval for Financial RAG: Resolving the Robustness-Precision Trade-offcited by 13
  9. 2026FASA: Frequency-aware Sparse Attentioncited by 8
  10. 2026VimRAG: Navigating Massive Visual Context in Retrieval-Augmented Generation via Multimodal Memory Graphcited by 5
  11. 2026SWE-Pruner Pro: The Coder LLM Already Knows What to Prunecited by 2
  12. 2026D2Plan: Dual-Agent Dynamic Global Planning for Complex Retrieval-Augmented Reasoningcited by 1
7 more
  1. 2026Mapping Text to Multiplex Graph: Prompt Compression as L'evy Walk-Guided Graph Pruningcited by 1
  2. 2026The Virtuous Cycle: AI-Powered Vector Search and Vector Search-Augmented AIcited by 0
  3. 2026Efficient Retrieval-Augmented Generation via Token Co-occurrence Graphscited by 0
  4. 2026Compression Beyond the Uncompressed: A Two-Stage Training Recipe for Soft Context Compression in RAGcited by 0
  5. 2026ARSM: Auto-Regressive State Machine for Agentic Reasoning Compressioncited by 0
  6. 2026EntroPrefill: Renyi-Guided Context Pruning with Conditional Stability Guarantees for Retrieval-Augmented Generationcited by 0
  7. 2026Clarify, Then Focus: Statement Normalization for Conversation Analytics at Scalecited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024MemoRAG: Boosting Long Context Processing with Global Memory-Enhanced Retrieval Augmentation
  2. 2025ParetoRAG: Leveraging Sentence-Context Attention for Robust and Efficient Retrieval-Augmented Generation
  3. 2025OpenRAG: Optimizing RAG End-to-End via In-Context Retrieval Learning
  4. 2026In-Context Optimization for Retrieval-Augmented Generation: A Gradient-Descent Perspective
  5. 2026Rethinking Retrieval-Augmentation as Synthesis: A Query-Aware Context Merging Approach
  6. 2025MacRAG: Compress, Slice, and Scale-up for Multi-Scale Adaptive Context RAG
  7. 2026AttentionRetriever: Attention Layers are Secretly Long Document Retrievers
  8. 2024PEAR: Position-Embedding-Agnostic Attention Re-weighting Enhances Retrieval-Augmented Generation with Zero Inference Overhead