Preventing posterior collapse in variational autoencoders for text generation via decoder regularization
arXiv:2110.14945
Abstract
Variational autoencoders trained to minimize the reconstruction error are sensitive to the posterior collapse problem, that is the proposal posterior distribution is always equal to the prior. We propose a novel regularization method based on fraternal dropout to prevent posterior collapse. We evaluate our approach using several metrics and observe improvements in all the tested configurations.
Accepted at NeurIPS 2021 Workshop DGMs Applications