Memorization in Overparameterized Autoencoders

Memorization of data in deep neural networks has become a subject of significant research interest. We prove that overparameterized single layer fully connected autoencoders memorize training data: they produce outputs in (a non-linear version of) the span of the training examples.   In contrast to fully connected autoencoders, we prove that depth is necessary for memorization in convolutional autoencoders.  Moreover, we observe that adding nonlinearity to deep convolutional autoencoders results in a stronger form of memorization: instead of outputting points in the span of the training images, deep convolutional autoencoders tend to output individual training images. Since convolutional autoencoder components are building blocks of deep convolutional networks, we envision that our findings will shed light on the important phenomenon of memorization in overparameterized deep networks.

Co-authors: Mikhail Belkin, Caroline Uhler