Multi-Agent Training beyond Zero-Sum with Correlated Equilibrium Meta-Solvers.

Luke Marris, Paul Muller, Marc Lanctot, Karl Tuyls, Thore Graepel

Browse the full ICML paper archive.