测量与增强LLM在RAG中的可信度：基于扎根归因与学会拒绝
Measuring and Enhancing Trustworthiness of LLMs in RAG through Grounded Attributions and Learning to Refuse
AI总结：
针对RAG中LLM可信度评估空白，提出Trust-Score指标与Trust-Align对齐方法，显著提升模型在ASQA、QAMPARI、ELI5上的表现及拒绝与引用能力。
摘要：
大型语言模型（LLMs）是检索增强生成（RAG）系统的重要组成部分。尽管许多研究聚焦于评估端到端RAG系统的整体质量，但在理解LLMs是否适合RAG任务方面仍存在空白。为解决此问题，我们引入了Trust-Score，一个在RAG框架内评估LLMs可信度的整体指标。我们的结果表明，各种提示方法，如上下文学习，未能有效使LLMs适应RAG任务（以Trust-Score衡量）。因此，我们提出了Trust-Align，一种对齐LLMs以提升Trust-Score表现的方法。使用Trust-Align对齐的27个模型中有26个在ASQA、QAMPARI和ELI5上大幅优于竞争基线。具体而言，在LLaMA-3-8b上，Trust-Align在ASQA（提升12.56）、QAMPARI（提升36.04）和ELI5（提升17.69）上优于FRONT。Trust-Align还显著增强了模型正确拒绝（弃权不执行）和提供高质量引用的能力。我们还展示了Trust-Align在不同开放权重模型上的有效性，包括LLaMA系列（1b至8b）、Qwen-2.5系列（0.5b至7b）和Phi3.5（3.8b）。我们在https://github.com/declare-lab/trust-align发布代码。
英文摘要：
LLMs are an integral component of retrieval-augmented generation (RAG) systems. While many studies focus on evaluating the overall quality of end-to-end RAG systems, there is a gap in understanding the appropriateness of LLMs for the RAG task. To address this, we introduce Trust-Score, a holistic metric that evaluates the trustworthiness of LLMs within the RAG framework. Our results show that various prompting methods, such as in-context learning, fail to effectively adapt LLMs to the RAG task as measured by Trust-Score. Consequently, we propose Trust-Align, a method to align LLMs for improved Trust-Score performance. 26 out of 27 models aligned using Trust-Align substantially outperform competitive baselines on ASQA, QAMPARI, and ELI5. Specifically, in LLaMA-3-8b, Trust-Align outperforms FRONT on ASQA (up 12.56), QAMPARI (up 36.04), and ELI5 (up 17.69). Trust-Align also significantly enhances models' ability to correctly refuse and provide quality citations. We also demonstrate the effectiveness of Trust-Align across different open-weight models, including the LLaMA series (1b to 8b), Qwen-2.5 series (0.5b to 7b), and Phi3.5 (3.8b). We release our code at https://github.com/declare-lab/trust-align.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
