Reviews: Hamiltonian Variational Auto-Encoder

Thank you for the thoughtful response --- I have read the response and the other author reviews. I believe the authors have addressed some of my concerns, although I still believe the empirical section could be more compelling. For instance, optimization tuning accounted for more of an improvement than the HVAE over the baseline. I realize this is a generally applicable concern, and I do like the idea and its presentation in this paper. I am maintaining my score of a 6. -------------- The authors present a new way to approximate the posterior distribution over latent variables in a deep generative model.