Normalizing Flow
A class of generative models that transforms a simple base probability distribution into a highly complex one using a sequence of invertible functions.
Think of It Like This
Like molding a perfectly round ball of clay into a complex statue in a way that you can precisely reverse back into a ball.
Because the transformations are strictly invertible, normalizing flows allow for exact, tractable likelihood estimation. This contrasts with GANs or VAEs, which only approximate the likelihood. They are powerful for density estimation and generating high-quality samples, though they require specific architectural constraints like easily computable Jacobians.