Higher-Order Uncoupled Dynamics Do Not Lead to Nash Equilibrium -- Except When They Do

Neural Information Processing Systems 

The framework of multi-agent learning explores the dynamics of how an agent's