Concept Whitening for Interpretable Image Recognition

What does a neural network encode about a concept as we traverse through the layers? Interpretability in machine learning is undoubtedly important, but the calculations of neural networks are very challenging to understand. Attempts to see inside their hidden layers can either be misleading, unusable, or rely on the latent space to possess properties that it may not have. In this work, rather than attempting to analyze a neural network posthoc, we introduce a mechanism, called concept whitening (CW), to alter a given layer of the network to allow us to better understand the computation leading up to that layer. When a concept whitening module is added to a CNN, the axes of the latent space can be aligned with concepts of interest. By experiment, we show that CW can provide us a much clearer understanding for how the network gradually learns concepts over layers without hurting predictive performance.