Auto-Encoding Explanatory Examples

César Ojeda, David Biesner, Ramses Sanchez, Kostadin Cvejoski, Jannis Schuecker, Christian Bauckhage, Bodgan Georgiev

Keywords: Variational Auto Encoders, Interpolations, Explanations, Stochastic Processes

TL;DR: We generate examples to explain a classifier desicion via interpolations in latent space. The variational auto encoder cost is extended with a functional of the classifier over the generated example path in data space.

Abstract:

In this paper, we ask for the main factors that determine a classifier's decision making and uncover such factors by studying latent codes produced by auto-encoding frameworks. To deliver an explanation of a classifier's behaviour, we propose a method that provides series of examples highlighting semantic differences between the classifier's decisions. We generate these examples through interpolations in latent space. We introduce and formalize the notion of a semantic stochastic path, as a suitable stochastic process defined in feature space via latent code interpolations. We then introduce the concept of semantic Lagrangians as a way to incorporate the desired classifier's behaviour and find that the solution of the associated variational problem allows for highlighting differences in the classifier decision.
Very importantly, within our framework the classifier is used as a black-box, and only its evaluation is required.

Code: https://figshare.com/s/91c74f38da02fb37877a