Don't Lie to Me! Robust and Efficient Explainability with Verified Perturbation Analysis

TL;DR

This paper introduces EVA, a novel explainability method for deep neural networks that guarantees exhaustive exploration of perturbation space, improving reliability and efficiency over existing sampling-based approaches.

Contribution

EVA is the first explainability technique to provide guaranteed complete coverage of perturbation space using verified perturbation analysis, enhancing explanation accuracy and computational efficiency.

Findings

Achieves state-of-the-art results on multiple benchmarks.

Provides exhaustive and reliable importance maps.

Outperforms sampling-based explainability methods.

Abstract

A variety of methods have been proposed to try to explain how deep neural networks make their decisions. Key to those approaches is the need to sample the pixel space efficiently in order to derive importance maps. However, it has been shown that the sampling methods used to date introduce biases and other artifacts, leading to inaccurate estimates of the importance of individual pixels and severely limit the reliability of current explainability methods. Unfortunately, the alternative -- to exhaustively sample the image space is computationally prohibitive. In this paper, we introduce EVA (Explaining using Verified perturbation Analysis) -- the first explainability method guarantee to have an exhaustive exploration of a perturbation space. Specifically, we leverage the beneficial properties of verified perturbation analysis -- time efficiency, tractability and guaranteed complete coverage of a manifold -- to efficiently characterize the input variables that are most likely to drive the model decision. We evaluate the approach systematically and demonstrate state-of-the-art results on multiple benchmarks.