Timezone: »

 
Poster
Turbocharging Solution Concepts: Solving NEs, CEs and CCEs with Neural Equilibrium Solvers
Luke Marris · Ian Gemp · Thomas Anthony · Andrea Tacchetti · Siqi Liu · Karl Tuyls

Thu Dec 01 02:00 PM -- 04:00 PM (PST) @ Hall J #718
in Poster Session 6 »

Solution concepts such as Nash Equilibria, Correlated Equilibria, and Coarse Correlated Equilibria are useful components for many multiagent machine learning algorithms. Unfortunately, solving a normal-form game could take prohibitive or non-deterministic time to converge, and could fail. We introduce the Neural Equilibrium Solver which utilizes a special equivariant neural network architecture to approximately solve the space of all games of fixed shape, buying speed and determinism. We define a flexible equilibrium selection framework, that is capable of uniquely selecting an equilibrium that minimizes relative entropy, or maximizes welfare. The network is trained without needing to generate any supervised training data. We show remarkable zero-shot generalization to larger games. We argue that such a network is a powerful component for many possible multiagent algorithms.

Keywords: Game Theory · Nash equilibrium · Coarse Correlated Equilibrium · Correlated Equilibrium
[ Paper]  [ Slides]  [ Poster]  [ OpenReview] 

Author Information

Luke Marris (DeepMind)
Ian Gemp (DeepMind)
Thomas Anthony (DeepMind)
Andrea Tacchetti (DeepMind)
Siqi Liu (Google DeepMind)
Karl Tuyls (DeepMind)

More from the Same Authors