Towards Best Practice in Explaining Neural Network Decisions with LRP

TL;DR

This paper evaluates the current best practices of Layer-wise Relevance Propagation (LRP) in explaining neural network decisions, demonstrating that layer-dependent application improves interpretability and object localization in visual detection tasks.

Contribution

It quantifies the impact of the best practice in applying LRP, confirming its effectiveness in better representing model reasoning and enhancing explanation quality.

Findings

Layer-dependent LRP better reflects neural network reasoning.

Improved object localization and class discriminativity with current LRP practices.

Quantitative evidence supporting the adoption of best practices in LRP application.

Abstract

Within the last decade, neural network based predictors have demonstrated impressive - and at times super-human - capabilities. This performance is often paid for with an intransparent prediction process and thus has sparked numerous contributions in the novel field of explainable artificial intelligence (XAI). In this paper, we focus on a popular and widely used method of XAI, the Layer-wise Relevance Propagation (LRP). Since its initial proposition LRP has evolved as a method, and a best practice for applying the method has tacitly emerged, based however on humanly observed evidence alone. In this paper we investigate - and for the first time quantify - the effect of this current best practice on feedforward neural networks in a visual object detection setting. The results verify that the layer-dependent approach to LRP applied in recent literature better represents the model's reasoning, and at the same time increases the object localization and class discriminativity of LRP.