How Inverse Conditional Flows Can Serve as a Substitute for Distributional Regression