Direct Optimization through $\arg \max$ for Discrete Variational Auto-Encoder

Sep 27, 2018 ICLR 2019 Conference Blind Submission readers: everyone Show Bibtex
  • Abstract: Reparameterization of variational auto-encoders is an effective method for reducing the variance of their gradient estimates. However, when the latent variables are discrete, a reparameterization is problematic due to discontinuities in the discrete space. In this work, we extend the direct loss minimization technique to discrete variational auto-encoders. We first reparameterize a discrete random variable using the $\arg \max$ function of the Gumbel-Max perturbation model. We then use direct optimization to propagate gradients through the non-differentiable $\arg \max$ using two perturbed $\arg \max$ operations.
  • Keywords: discrete variational auto encoders, generative models, perturbation models
0 Replies

Loading