We propose a novel regularization loss that enforces standard Gaussianity, encouraging samples to align with a standard Gaussian distribution.