no code implementations • 20 Sep 2022 • Hanao Li, Tian Han
In this paper, we present a new unsupervised learning method to enforce sparsity on the latent space for the generator model with a gradually sparsified spike and slab distribution as our prior.