What specific divergence measure is used in Variational Autoencoders (VAEs) to ensure the learned latent distribution stays close to a prior distribution?
Show answer
Kullback-Leibler divergence
The KL divergence acts as a regularization term in the VAE loss function. It measures how one probability distribution differs from a second, reference probability distribution, typically a standard normal distribution N(0,1).
Also asked as:
- In Variational Autoencoders, what divergence measure is minimized to keep the latent distribution similar to the prior?