Interpretability of Deep Learning (DL) is a barrier to trustworthy AI. Despite great efforts made by the Explainable AI (XAI) community, explanations lack robustness -- indistinguishable input perturbations may lead to different XAI results. Thus, it is vital to assess how robust DL interpretability is, given an XAI method. In this paper, we identify several challenges that the state-of-the-art is unable to cope with collectively: i) existing metrics are not comprehensive; ii) XAI techniques are highly heterogeneous; iii) misinterpretations are normally rare events. To tackle these challenges, we introduce two black-box evaluation methods, concerning the worst-case interpretation discrepancy and a probabilistic notion of how robust in general, respectively. Genetic Algorithm (GA) with bespoke fitness function is used to solve constrained optimisation for efficient worst-case evaluation. Subset Simulation (SS), dedicated to estimate rare event probabilities, is used for evaluating overall robustness. Experiments show that the accuracy, sensitivity, and efficiency of our methods outperform the state-of-the-arts. Finally, we demonstrate two applications of our methods: ranking robust XAI methods and selecting training schemes to improve both classification and interpretation robustness.

翻译：深度学习（DL）的可解释性是通往可信人工智能的障碍。尽管可解释人工智能（XAI）领域付出了巨大努力，但解释缺乏鲁棒性——难以区分的输入扰动可能导致不同的XAI结果。因此，在给定XAI方法的情况下，评估DL可解释性的鲁棒性至关重要。本文指出当前先进方法无法共同应对的若干挑战：i) 现有指标不全面；ii) XAI技术高度异质；iii) 误解通常属于罕见事件。为应对这些挑战，我们提出两种黑盒评估方法：分别针对最坏情况下的解释差异和鲁棒性的概率概念（即整体鲁棒性）。采用定制适应度函数的遗传算法（GA）求解约束优化问题，以实现高效的最坏情况评估；采用专用于估计罕见事件概率的子集模拟（SS）评估整体鲁棒性。实验表明，我们方法的准确性、敏感性和效率均优于现有方法。最后，我们演示了这两种方法的两类应用：对鲁棒XAI方法进行排序，以及选择训练方案以同时提升分类鲁棒性和解释鲁棒性。