Hamiltonian Monte Carlo on ReLU Neural Networks is Inefficient

Open in new window