Lune

ACL Findings2026

PRA-RAG: Provably Robust Aggregation in Retrieval-Augmented Generation against Retrieval Corruption

Xue Tan, Yi Zheng, Chang Huo, Yunruo Zhang, Yu Liu, Hao Luan, Zhuyang Yu, Jun Dai, Xiaoyan Sun, Ping Chen

Abstract

Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by incorporating external knowledge, effectively mitigating their inherent knowledge limitations. However, RAG remains vulnerable to poisoning attacks that manipulate retrieved texts to mislead model outputs. Existing defense mechanisms often lack theoretical robustness guarantees and perform unreliably when the LLM has limited knowledge of the retrieved content. In this work, we propose PRA-RAG, a provably robust retrieval aggregation algorithm designed to defend against poisoning attacks on retrieved texts. PRA-RAG samples multiple combinations of retrieved texts and utilizes geometric structures in the embedding space to identify a robust subset, from which a stable aggregated representation is derived. We provide theoretical bounds on the maximum impact of poisoned retrieved content and establish a quantitative measure of RAG's robustness. Experiments across multiple benchmarks and RAG architectures demonstrate that PRA-RAG reduces the attack success rate to as low as 1% while maintaining an accuracy of 71%, significantly outperforming representative state-ofthe-art (SOTA) methods.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Making Retrieval-Augmented Language Models Robust to Irrelevant Context cites Measuring and Narrowing the Compositionality Gap in Language Models
  • Making Retrieval-Augmented Language Models Robust to Irrelevant Context cites Language Models are Few-Shot Learners
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on13

  1. Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al.NeurIPS 2020 · 64,255 citations
  2. Self-RAG: Learning to Retrieve, Generate, and Critique through Self-ReflectionAkari Asai, Zeqiu Wu, Yizhong Wang, Avirup Sil et al.ICLR 2024 · 1,798 citations
  3. Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text RetrievalLee Xiong, Chenyan Xiong, Ye Li, Kwok-Fung Tang et al.ICLR 2021 · 1,547 citations
  4. Making Retrieval-Augmented Language Models Robust to Irrelevant ContextOri Yoran, Tomer Wolfson, Ori Ram, Jonathan BerantICLR 2024 · 361 citations
  5. Measuring and Narrowing the Compositionality Gap in Language ModelsOfir Press, Muru Zhang, Sewon Min, Ludwig Schmidt et al.EMNLP Findings 2023 · 296 citations

Related papers

  1. IRAG: Robust Multimodal Retrieval-Augmented Generation via Hazard SeparationRuikun Luo, Zixiao Feng, Lin Gu, Xiaoyu XiaWWW 2026
  2. On the Vulnerability of Applying Retrieval-Augmented Generation within Knowledge-Intensive Application DomainsXun Xian, Ganghua Wang, Xuan Bi, Rui Zhang et al.ICML 2025
  3. ReliabilityRAG: Effective and Provably Robust Defense for RAG-based Web-SearchZeyu Shen, Basileal Imana, Tong Wu, Chong Xiang et al.NeurIPS 2025 · 26 citations
  4. Reranker Helps, but Not Enough: Towards Strong Poisoning Attacks Against Retrieval-Augmented GenerationXiaokun Yang, Jian Liang, Yesheng Liu, Xin Xiong et al.ICML 2026
  5. Traceback of Poisoning Attacks to Retrieval-Augmented GenerationBaolei Zhang, Haoran Xin, Minghong Fang, Zhuqing Liu et al.WWW 2025 · 20 citations
Dusk over the sea between two cliffs drawn in fine vertical lines