On Network Science and Mutual Information for Explaining Deep Neural Networks
This work addresses the challenge of explaining deep neural networks for researchers and practitioners, but it appears incremental as it builds on existing concepts of mutual information and network science.
The paper tackles the problem of interpreting deep learning models by proposing a new approach that combines mutual information and network science to quantify information flow between neurons, resulting in a technique called Neural Information Flow (NIF) that provides feature attributions.
In this paper, we present a new approach to interpret deep learning models. By coupling mutual information with network science, we explore how information flows through feedforward networks. We show that efficiently approximating mutual information allows us to create an information measure that quantifies how much information flows between any two neurons of a deep learning model. To that end, we propose NIF, Neural Information Flow, a technique for codifying information flow that exposes deep learning model internals and provides feature attributions.