Variational Autoencoders (2013)

Kingma and Welling (2013) combined autoencoders with variational inference, creating a principled framework for learning latent representations and generating new samples.