Hamiltonian Monte Carlo has a beautiful idea at its core: instead of proposing random steps and hoping for the best, let the gradient of your log-density guide the sampler like...