Maximum Entropy Baseline for Integrated Gradients

Hanxiao Tan

Abstract: Integrated Gradients (IG), one of the most popular explainability methods available, still remains ambiguous in the selection of baseline, which may seriously impair the credibility of the explanations. This study proposes a new uniform baseline, i.e., the Maximum Entropy Baseline, which is consistent with the “uninformative” property of baselines defined in IG. In addition, we propose an improved ablating evaluation approach incorporating the new baseline, where the information conservativeness is maintained. Furthermore, we explain the linear transformation invariance of IG baselines from an information perspective. Finally, extensive experiments indicate that IG with Maximum Entropy Baseline performs superiorly and the ablation tests derived from it are more plausible.