XplainLLM

XplainLLM是一个专为理解大型语言模型（LLM）决策过程而设计的数据集，由加州大学圣巴巴拉分校创建。该数据集包含12,102个问答解释（QAE）三元组，每个解释都链接了LLM的推理过程与知识图谱（KG）中的实体和关系。通过集成KG和图注意力网络（GAT），XplainLLM提供了对LLM决策过程的人类可理解解释，旨在提高LLM的透明度和可靠性，适用于增强模型解释性和提升模型性能的研究领域。

XplainLLM is a dataset specifically designed to understand the decision-making processes of Large Language Models (LLMs), which was created by the University of California, Santa Barbara. The dataset contains 12,102 question-answer-explanation (QAE) triples, where each explanation links the reasoning process of an LLM to entities and relationships in a Knowledge Graph (KG). By integrating Knowledge Graphs (KGs) and Graph Attention Networks (GATs), XplainLLM provides human-interpretable explanations for the decision-making processes of LLMs, aiming to improve the transparency and reliability of LLMs and is applicable to research fields focused on enhancing model interpretability and improving model performance.