Inducing Interpretable Representations with Variational Autoencoders
This work addresses the challenge of interpretability in deep learning for researchers and practitioners, but it appears incremental as it builds on existing VAE frameworks.
The authors tackled the problem of learning interpretable representations in variational autoencoders by incorporating structured graphical models into the encoders, enabling reasoning under structural constraints and handling high-dimensional data with deep generative models.
We develop a framework for incorporating structured graphical models in the \emph{encoders} of variational autoencoders (VAEs) that allows us to induce interpretable representations through approximate variational inference. This allows us to both perform reasoning (e.g. classification) under the structural constraints of a given graphical model, and use deep generative models to deal with messy, high-dimensional domains where it is often difficult to model all the variation. Learning in this framework is carried out end-to-end with a variational objective, applying to both unsupervised and semi-supervised schemes.