Our paper, RW-Flow, which introduces a one-step generative model for learning distributions on compact manifolds via Wasserstein gradient flows, is now available on arXiv.