CrAM: Credibility-Aware Attention Modification in LLMs for Combating Misinformation in RAG

• URL: http://arxiv.org/abs/2406.11497v2
• Date: Thu, 27 Jun 2024 10:18:53 GMT
• Title: CrAM: Credibility-Aware Attention Modification in LLMs for Combating Misinformation in RAG
• Authors: Boyi Deng, Wenjie Wang, Fengbin Zhu, Qifan Wang, Fuli Feng,
• Abstract summary: Retrieval-Augmented Generation (RAG) can alleviate hallucinations of Large Language Models (LLMs) by referencing external documents. To address this issue, we explore the task of "credibility-aware RAG" We introduce a plug-and-play method named $textbfCr$edibility-aware $textbfA$ttention $textbfM$odification (CrAM) CrAM identifies influential attention heads and adjusts their attention weights based on the credibility of the documents, thereby reducing the impact of low-credibility documents.
• Score: 50.030526904378256
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Retrieval-Augmented Generation (RAG) can alleviate hallucinations of Large Language Models (LLMs) by referencing external documents. However, the misinformation in external documents may mislead LLMs' generation. To address this issue, we explore the task of "credibility-aware RAG", in which LLMs automatically adjust the influence of retrieved documents based on their credibility scores to counteract misinformation. To this end, we introduce a plug-and-play method named $\textbf{Cr}$edibility-aware $\textbf{A}$ttention $\textbf{M}$odification (CrAM). CrAM identifies influential attention heads in LLMs and adjusts their attention weights based on the credibility of the documents, thereby reducing the impact of low-credibility documents. Experiments on Natual Questions and TriviaQA using Llama2-13B, Llama3-8B, and Qwen-7B show that CrAM improves the RAG performance of LLMs against misinformation pollution by over 20%, even surpassing supervised fine-tuning methods.

Related papers

• Invar-RAG: Invariant LLM-aligned Retrieval for Better Generation [43.630437906898635]
  We propose a novel two-stage fine-tuning architecture called Invar-RAG. In the retrieval stage, an LLM-based retriever is constructed by integrating LoRA-based representation learning. In the generation stage, a refined fine-tuning method is employed to improve LLM accuracy in generating answers based on retrieved information.
  arXiv  Detail & Related papers  (2024-11-11T14:25:37Z)
• Evaluation of Attribution Bias in Retrieval-Augmented Large Language Models [47.694137341509304]
  We evaluate the attribution sensitivity and bias with respect to authorship information in large language models. Our results show that adding authorship information to source documents can significantly change the attribution quality of LLMs by 3% to 18%. Our findings indicate that metadata of source documents can influence LLMs' trust, and how they attribute their answers.
  arXiv  Detail & Related papers  (2024-10-16T08:55:49Z)
• LLM Self-Correction with DeCRIM: Decompose, Critique, and Refine for Enhanced Following of Instructions with Multiple Constraints [86.59857711385833]
  We introduce RealInstruct, the first benchmark designed to evaluate LLMs' ability to follow real-world multi-constrained instructions. To address the performance gap between open-source and proprietary models, we propose the Decompose, Critique and Refine (DeCRIM) self-correction pipeline. Our results show that DeCRIM improves Mistral's performance by 7.3% on RealInstruct and 8.0% on IFEval even with weak feedback.
  arXiv  Detail & Related papers  (2024-10-09T01:25:10Z)
• MQM-APE: Toward High-Quality Error Annotation Predictors with Automatic Post-Editing in LLM Translation Evaluators [53.91199933655421]
  Large Language Models (LLMs) have shown significant potential as judges for Machine Translation (MT) quality assessment. We introduce a universal and training-free framework, $textbfMQM-APE, to enhance the quality of error annotations predicted by LLM evaluators.
  arXiv  Detail & Related papers  (2024-09-22T06:43:40Z)
• Refiner: Restructure Retrieval Content Efficiently to Advance Question-Answering Capabilities [30.1331670544648]
  Large Language Models (LLMs) are limited by their parametric knowledge, leading to hallucinations in knowledge-extensive tasks. We propose $textitRefiner$, an end-to-end extract-and-restructure paradigm that operates in the post-retrieval process of RAG.
  arXiv  Detail & Related papers  (2024-06-17T09:25:10Z)
• Unsupervised Information Refinement Training of Large Language Models for Retrieval-Augmented Generation [128.01050030936028]
  We propose an information refinement training method named InFO-RAG. InFO-RAG is low-cost and general across various tasks. It improves the performance of LLaMA2 by an average of 9.39% relative points.
  arXiv  Detail & Related papers  (2024-02-28T08:24:38Z)
• GenAudit: Fixing Factual Errors in Language Model Outputs with Evidence [64.95492752484171]
  We present GenAudit -- a tool intended to assist fact-checking LLM responses for document-grounded tasks. We train models to execute these tasks, and design an interactive interface to present suggested edits and evidence to users. To ensure that most errors are flagged by the system, we propose a method that can increase the error recall while minimizing impact on precision.
  arXiv  Detail & Related papers  (2024-02-19T21:45:55Z)
• LLatrieval: LLM-Verified Retrieval for Verifiable Generation [67.93134176912477]
  Verifiable generation aims to let the large language model (LLM) generate text with supporting documents. We propose LLatrieval (Large Language Model Verified Retrieval), where the LLM updates the retrieval result until it verifies that the retrieved documents can sufficiently support answering the question. Experiments show that LLatrieval significantly outperforms extensive baselines and achieves state-of-the-art results.
  arXiv  Detail & Related papers  (2023-11-14T01:38:02Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.