论文标题
SAM：归因方法对超参数的敏感性
SAM: The Sensitivity of Attribution Methods to Hyperparameters
论文作者
论文摘要
归因方法可以为分类器决定的原因提供有力的见解。我们认为，解释方法的关键避免是其对输入超参数的鲁棒性，通常是随机设置或经验调整的。对任意超参数选择的高度敏感性不仅会阻碍可重复性，而且质疑解释的正确性并损害了最终用户的信任。在本文中，我们提供了有关现有归因方法敏感性的彻底实证研究。我们发现，令人震惊的趋势是，许多方法对它们常见的超参数（例如它们）的变化高度敏感。即使改变随机种子也会产生不同的解释！有趣的是，这种敏感性并未反映在文献中通常报道的数据集的平均解释精度得分中。此外，与适用于普通分类器生成的分类器生成的解释（即对像素扰动不变的训练为不变性的训练）更强大。
Attribution methods can provide powerful insights into the reasons for a classifier's decision. We argue that a key desideratum of an explanation method is its robustness to input hyperparameters which are often randomly set or empirically tuned. High sensitivity to arbitrary hyperparameter choices does not only impede reproducibility but also questions the correctness of an explanation and impairs the trust of end-users. In this paper, we provide a thorough empirical study on the sensitivity of existing attribution methods. We found an alarming trend that many methods are highly sensitive to changes in their common hyperparameters e.g. even changing a random seed can yield a different explanation! Interestingly, such sensitivity is not reflected in the average explanation accuracy scores over the dataset as commonly reported in the literature. In addition, explanations generated for robust classifiers (i.e. which are trained to be invariant to pixel-wise perturbations) are surprisingly more robust than those generated for regular classifiers.
