generative adversarial networks (GANs) can successfully approximate a
probability distribution and produce realistic samples. However, open questions
such as sufficient convergence conditions and mode collapse still persist. In
this paper, we build on existing work in the area by propo