Timezone: »

A mean-field analysis of two-player zero-sum games
Carles Domingo-Enrich · Samy Jelassi · Arthur Mensch · Grant Rotskoff · Joan Bruna

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1638

Finding Nash equilibria in two-player zero-sum continuous games is a central problem in machine learning, e.g. for training both GANs and robust models. The existence of pure Nash equilibria requires strong conditions which are not typically met in practice. Mixed Nash equilibria exist in greater generality and may be found using mirror descent. Yet this approach does not scale to high dimensions. To address this limitation, we parametrize mixed strategies as mixtures of particles, whose positions and weights are updated using gradient descent-ascent. We study this dynamics as an interacting gradient flow over measure spaces endowed with the Wasserstein-Fisher-Rao metric. We establish global convergence to an approximate equilibrium for the related Langevin gradient-ascent dynamic. We prove a law of large numbers that relates particle dynamics to mean-field dynamics. Our method identifies mixed equilibria in high dimensions and is demonstrably effective for training mixtures of GANs.

Author Information

Carles Domingo-Enrich (NYU)
Samy Jelassi (Princeton University)
Arthur Mensch (ENS)
Grant Rotskoff (Stanford University)
Joan Bruna (NYU)

More from the Same Authors