I think the two methods generate the output from the same distribution. Below method exploits "from torch.distributions import MultivariateNormal, Normal, Independe