Explaining Classifiers by Constructing Familiar Concepts
This addresses the challenge of making deep learning models more interpretable for users, though it is incremental as it builds on existing decoder-based methods.
The paper tackles the problem of interpreting neurons in deep neural networks by proposing ClaDec, an architecture that decodes classifier layers into human-familiar representations, showing in evaluations that it captures more classification-relevant information than auto-encoders and highlights more relevant input areas than saliency methods.
Interpreting a large number of neurons in deep learning is difficult. Our proposed `CLAssifier-DECoder' architecture (ClaDec) facilitates the understanding of the output of an arbitrary layer of neurons or subsets thereof. It uses a decoder that transforms the incomprehensible representation of the given neurons to a representation that is more similar to the domain a human is familiar with. In an image recognition problem, one can recognize what information (or concepts) a layer maintains by contrasting reconstructed images of ClaDec with those of a conventional auto-encoder(AE) serving as reference. An extension of ClaDec allows trading comprehensibility and fidelity. We evaluate our approach for image classification using convolutional neural networks. We show that reconstructed visualizations using encodings from a classifier capture more relevant classification information than conventional AEs. This holds although AEs contain more information on the original input. Our user study highlights that even non-experts can identify a diverse set of concepts contained in images that are relevant (or irrelevant) for the classifier. We also compare against saliency based methods that focus on pixel relevance rather than concepts. We show that ClaDec tends to highlight more relevant input areas to classification though outcomes depend on classifier architecture. Code is at \url{https://github.com/JohnTailor/ClaDec}