2022 (modified: 17 Nov 2022)ICLR 2022Readers: Everyone
Abstract:The recent success of autoencoders for representation learning can be traced in large part to the addition of a regularization term. Such regularized autoencoders ``constrain" the representation so...