Training a Hopfield Variational Autoencoder with Equilibrium Propagation
This work addresses the need for more efficient analog hardware for generative AI, though it is incremental in extending equilibrium propagation from discriminative to generative settings.
The paper tackled the problem of applying equilibrium propagation to generative modeling by training a variational autoencoder (VAE) with a Hopfield network, resulting in a design that could halve the required chip size for analog hardware implementations.
On dedicated analog hardware, equilibrium propagation is an energy-efficient alternative to backpropagation. In spite of its theoretical guarantees, its application in the AI domain remains limited to the discriminative setting. Meanwhile, despite its high computational demands, generative AI is on the rise. In this paper, we demonstrate the application of Equilibrium Propagation in training a variational autoencoder (VAE) for generative modeling. Leveraging the symmetric nature of Hopfield networks, we propose using a single model to serve as both the encoder and decoder which could effectively halve the required chip size for VAE implementations, paving the way for more efficient analog hardware configurations.