# VAE

An autoencoder that learns a distribution over a latent space and is trained with a variational objective.

## What it is used for

Generative modelling, anomaly detection by reconstruction, and latent representations.