Hamiltonian evolution for the hybrid Monte Carlo algorithm

We discuss a class of reversible, discrete approximations to Hamilton's equations which can be used to improve the efficiency of the hybrid Monte Carlo algorithm. © 1992.