# ClashEval: Quantifying the tug-of-war between an LLM's internal prior and external evidence

arXiv: https://arxiv.org/abs/2404.10198 · HTML: https://mlsanity.com/paper/2404.10198/

## Key facts

- Cited by 113 papers in the map (data through 14 Oct 2026).
- Placed in RAG robustness and adaptive retrieval, one of 15 directions in Retrieval-augmented generation.
- Builds on 20 of its 28 references that are in the map.
- First posted on arXiv 16 Apr 2024.

## Placed in

- Retrieval-augmented generation / [RAG robustness and adaptive retrieval](https://mlsanity.com/rag/rag-robustness-and-adaptive-retrieval/)

## Builds on (most cited first, 20 of 20)

- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022)
- [DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature](https://mlsanity.com/paper/2301.11305/) (2023)
- [Retrieval Augmentation Reduces Hallucination in Conversation](https://mlsanity.com/paper/2104.07567/) (2021)
- [Ragas: Automated Evaluation of Retrieval Augmented Generation](https://mlsanity.com/paper/2309.15217/) (2023)
- [Entity-Based Knowledge Conflicts in Question Answering](https://mlsanity.com/paper/2109.05052/) (2021)
- [Adaptive Chameleon or Stubborn Sloth: Revealing the Behavior of Large Language Models in Knowledge Conflicts](https://mlsanity.com/paper/2305.13300/) (2023)
- [Benchmarking Large Language Models in Retrieval-Augmented Generation](https://mlsanity.com/paper/2309.01431/) (2023)
- [TrustLLM: Trustworthiness in Large Language Models](https://mlsanity.com/paper/2401.05561/) (2024)
- [ARES: An Automated Evaluation Framework for Retrieval-Augmented Generation Systems](https://mlsanity.com/paper/2311.09476/) (2023)
- [Challenges and Applications of Large Language Models](https://mlsanity.com/paper/2307.10169/) (2023)
- [Generation-Augmented Retrieval for Open-domain Question Answering](https://mlsanity.com/paper/2009.08553/) (2020)
- [Med-HALT: Medical Domain Hallucination Test for Large Language Models](https://mlsanity.com/paper/2307.15343/) (2023)
- [Same Task, More Tokens: the Impact of Input Length on the Reasoning Performance of Large Language Models](https://mlsanity.com/paper/2402.14848/) (2024)
- [Ever: Mitigating Hallucination in Large Language Models through Real-Time Verification and Rectification](https://mlsanity.com/paper/2311.09114/) (2023)
- [Creating Trustworthy LLMs: Dealing with Hallucinations in Healthcare AI](https://mlsanity.com/paper/2311.01463/) (2023)
- [Evaluation of GPT-3.5 and GPT-4 for supporting real-world information needs in healthcare delivery](https://mlsanity.com/paper/2304.13714/) (2023)
- [RetrievalQA: Assessing Adaptive Retrieval-Augmented Generation for Short-form Open-Domain Question Answering](https://mlsanity.com/paper/2402.16457/) (2024)
- [RAGged Edges: The Double-Edged Sword of Retrieval-Augmented Chatbots](https://mlsanity.com/paper/2403.01193/) (2024)
- [RaLLe: A Framework for Developing and Evaluating Retrieval-Augmented Large Language Models](https://mlsanity.com/paper/2308.10633/) (2023)

## Built on by (most cited first, 30 of 113)

- [Words or Vision: Do Vision-Language Models Have Blind Faith in Text?](https://mlsanity.com/paper/2503.02199/) (2025)
- [FaithEval: Can Your Language Model Stay Faithful to Context, Even If "The Moon is Made of Marshmallows"](https://mlsanity.com/paper/2410.03727/) (2024)
- [Rationale-Guided Retrieval Augmented Generation for Medical Question Answering](https://mlsanity.com/paper/2411.00300/) (2024)
- [RAG-DDR: Optimizing Retrieval-Augmented Generation Using Differentiable Data Rewards](https://mlsanity.com/paper/2410.13509/) (2024)
- [SafeRAG: Benchmarking Security in Retrieval-Augmented Generation of Large Language Model](https://mlsanity.com/paper/2501.18636/) (2025)
- [Beyond Text: Optimizing RAG with Multimodal Inputs for Industrial Applications](https://mlsanity.com/paper/2410.21943/) (2024)
- [To Trust or Not to Trust? Enhancing Large Language Models' Situated Faithfulness to External Contexts](https://mlsanity.com/paper/2410.14675/) (2024)
- [RAG LLMs are Not Safer: A Safety Analysis of Retrieval-Augmented Generation for Large Language Models](https://mlsanity.com/paper/2504.18041/) (2025)
- [Enhancing Retrieval-Augmented Generation: A Study of Best Practices](https://mlsanity.com/paper/2501.07391/) (2025)
- [Grounding and Evaluation for Large Language Models: Practical Challenges and Lessons Learned (Survey)](https://mlsanity.com/paper/2407.12858/) (2024)
- [KnowPO: Knowledge-aware Preference Optimization for Controllable Knowledge Selection in Retrieval-Augmented Language Models](https://mlsanity.com/paper/2408.03297/) (2024)
- [Teaching Models to Balance Resisting and Accepting Persuasion](https://mlsanity.com/paper/2410.14596/) (2024)
- [Compressing Long Context for Enhancing RAG with AMR-based Concept Distillation](https://mlsanity.com/paper/2405.03085/) (2024)
- [On the Intrinsic Self-Correction Capability of LLMs: Uncertainty and Latent Concept](https://mlsanity.com/paper/2406.02378/) (2024)
- [MEBench: Benchmarking Large Language Models for Cross-Document Multi-Entity Question Answering](https://mlsanity.com/paper/2502.18993/) (2025)
- [Shifting from Ranking to Set Selection for Retrieval Augmented Generation](https://mlsanity.com/paper/2507.06838/) (2025)
- [Does RAG Know When Retrieval Is Wrong? Diagnosing Context Compliance under Knowledge Conflict](https://mlsanity.com/paper/2605.14473/) (2026)
- [Hierarchical Retrieval-Augmented Generation Model with Rethink for Multi-hop Question Answering](https://mlsanity.com/paper/2408.11875/) (2024)
- [Parenting: Optimizing Knowledge Selection of Retrieval-Augmented Language Models with Parameter Decoupling and Tailored Tuning](https://mlsanity.com/paper/2410.10360/) (2024)
- [Don't lie to your friends: Learning what you know from collaborative self-play](https://mlsanity.com/paper/2503.14481/) (2025)
- [FineTuneBench: How well do commercial fine-tuning APIs infuse knowledge into LLMs?](https://mlsanity.com/paper/2411.05059/) (2024)
- [Feedback Friction: LLMs Struggle to Fully Incorporate External Feedback](https://mlsanity.com/paper/2506.11930/) (2025)
- [LLM Microscope: What Model Internals Reveal About Answer Correctness and Context Utilization](https://mlsanity.com/paper/2510.04013/) (2025)
- [Deciphering the Interplay of Parametric and Non-parametric Memory in Retrieval-augmented Language Models](https://mlsanity.com/paper/2410.05162/) (2024)
- [Quantifying and Improving the Robustness of Retrieval-Augmented Language Models Against Spurious Features in Grounding Data](https://mlsanity.com/paper/2503.05587/) (2025)
- [Medical large language models are easily distracted](https://mlsanity.com/paper/2504.01201/) (2025)
- [MultiFinRAG: An Optimized Multimodal Retrieval-Augmented Generation (RAG) Framework for Financial Question Answering](https://mlsanity.com/paper/2506.20821/) (2025)
- [Know Your RAG: Dataset Taxonomy and Generation Strategies for Evaluating RAG Systems](https://mlsanity.com/paper/2411.19710/) (2024)
- [Can Large Language Models Match the Conclusions of Systematic Reviews?](https://mlsanity.com/paper/2505.22787/) (2025)
- [SUGAR: Leveraging Contextual Confidence for Smarter Retrieval](https://mlsanity.com/paper/2501.04899/) (2025)

## Abstract (arXiv)

Retrieval augmented generation (RAG) is frequently used to mitigate hallucinations and provide up-to-date knowledge for large language models (LLMs). However, given that document retrieval is an imprecise task and sometimes results in erroneous or even harmful content being presented in context, this raises the question of how LLMs handle retrieved information: If the provided content is incorrect, does the model know to ignore it, or does it recapitulate the error? Conversely, when the model's initial response is incorrect, does it always know to use the retrieved information to correct itself, or does it insist on its wrong prior response? To answer this, we curate a dataset of over 1200 questions across six domains (e.g., drug dosages, Olympic records, locations) along with content relevant to answering each question. We further apply precise perturbations to the answers in the content that range from subtle to blatant errors. We benchmark six top-performing LLMs, including GPT-4o, on this dataset and find that LLMs are susceptible to adopting incorrect retrieved content, overriding their own correct prior knowledge over 60% of the time. However, the more unrealistic the retrieved content is (i.e. more deviated from truth), the less likely the model is to adopt it. Also, the less confident a model is in its initial response (via measuring token probabilities), the more likely it is to adopt the information in the retrieved content. We exploit this finding and demonstrate simple methods for improving model accuracy where there is conflicting retrieved content. Our results highlight a difficult task and benchmark for LLMs -- namely, their ability to correctly discern when it is wrong in light of correct retrieved content and to reject cases when the provided content is incorrect.