no code implementations • 19 Jul 2021 • Catherine F. Higham, Adrian Bedford
We demonstrate the feasibility of framing a classically learned deep neural network as an energy based model that can be processed on a one-step quantum annealer in order to exploit fast sampling times.