GCI：一个（G）概念（C）概念（I）解释框架
GCI: A (G)raph (C)oncept (I)nterpretation Framework
摘要：
可解释人工智能（XAI）最近在概念提取方面的研究激增，重点是从深度神经网络中提取人类可解释的概念。概念提取方法面临的一个重要挑战是难以解释和评估发现的概念，特别是对于复杂的任务，如分子性质预测。我们通过提出GCI来应对这一挑战：一个（G）概念（C）概念（I）解释框架，用于定量测量从图神经网络（GNN）发现的概念与其相应的人类解释之间的一致性。GCI将概念解释编码为函数，可用于定量测量给定解释与概念定义之间的一致性。我们展示了GCI的四个应用：（i）定量评估概念提取器，（ii）测量概念提取器和人类解释之间的对齐，（iii）测量关于最终任务的解释的完整性，以及（iv）GCI在分子性质预测中的实际应用，其中我们演示了如何使用化学官能团来解释在分子性质预测任务上训练的GNN，并以0.76的AUCROC完整性评分进行解释。
英文摘要：
Explainable AI (XAI) underwent a recent surge in research on concept extraction, focusing on extracting human-interpretable concepts from Deep Neural Networks. An important challenge facing concept extraction approaches is the difficulty of interpreting and evaluating discovered concepts, especially for complex tasks such as molecular property prediction. We address this challenge by presenting GCI: a (G)raph (C)oncept (I)nterpretation framework, used for quantitatively measuring alignment between concepts discovered from Graph Neural Networks (GNNs) and their corresponding human interpretations. GCI encodes concept interpretations as functions, which can be used to quantitatively measure the alignment between a given interpretation and concept definition. We demonstrate four applications of GCI: (i) quantitatively evaluating concept extractors, (ii) measuring alignment between concept extractors and human interpretations, (iii) measuring the completeness of interpretations with respect to an end task and (iv) a practical application of GCI to molecular property prediction, in which we demonstrate how to use chemical functional groups to explain GNNs trained on molecular property prediction tasks, and implement interpretations with a 0.76 AUCROC completeness score.
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。
