Optimizing Explanations by Network Canonization and Hyperparameter Search

Frederik Pahde, Galip Ümit Yolcu, Alexander Binder, Wojciech Samek, Sebastian Lapuschkin

Abstract:

Explainable AI (XAI) is slowly becoming a key component for many AI applications. Rule-based and modified
backpropagation XAI approaches however often face challenges when being applied to modern model architectures
including innovative layer building blocks, which is caused
by two reasons. Firstly, the high flexibility of rule-based XAI
methods leads to numerous potential parameterizations.
Secondly, many XAI methods break the implementationinvariance axiom because they struggle with certain model
components, e.g., BatchNorm layers. The latter can be addressed with model canonization, which is the process of restructuring the model to disregard problematic components
without changing the underlying function. While model
canonization is straightforward for simple architectures
(e.g., VGG, ResNet), it can be challenging for more complex
and highly interconnected models (e.g., DenseNet). Moreover, there is only little quantifiable evidence that model
canonization is beneficial for XAI. In this work, we propose
canonizations for currently relevant model blocks applicable to popular deep neural network architectures, including
VGG, ResNet, EfficientNet, DenseNets, as well as Relation
Networks. We further suggest a XAI evaluation framework
with which we quantify and compare the effects of model
canonization for various XAI methods in image classification tasks on the Pascal VOC and ILSVRC2017 datasets,
as well as for Visual Question Answering using CLEVRXAI. Moreover, addressing the former issue outlined above,
we demonstrate how our evaluation framework can be applied to perform hyperparameter search for XAI methods
to optimize the quality of explanations. Code is available
on https://github.com/frederikpahde/xaicanonization