RAGE Against the Machine:检索增强的大语言模型解释
RAGE Against the Machine: Retrieval-Augmented LLM Explanations
AI总结:
本文提出RAGE,一个交互式工具,通过反事实分析和剪枝方法解释检索增强的大语言模型,帮助用户识别影响答案的关键上下文并查看答案来源。
摘要:
本文展示了RAGE,一个用于解释具有检索能力的大语言模型(LLMs)的交互式工具;即能够查询外部来源并将相关信息拉入其输入上下文。我们的解释是反事实的,即它们识别输入上下文中在移除后会改变对LLM所提问题答案的部分。RAGE包括剪枝方法以导航可能的解释的广阔空间,允许用户查看所产生答案的来源。
英文摘要:
This paper demonstrates RAGE, an interactive tool for explaining Large Language Models (LLMs) augmented with retrieval capabilities; i.e., able to query external sources and pull relevant information into their input context. Our explanations are counterfactual in the sense that they identify parts of the input context that, when removed, change the answer to the question posed to the LLM. RAGE includes pruning methods to navigate the vast space of possible explanations, allowing users to view the provenance of the produced answers.
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。