A Unified Perspective on Deep Equilibrium Finding

TL;DR

This paper introduces a unified framework for deep equilibrium-finding algorithms in extensive-form games, combining and generalizing PSRO and CFR through novel neural modules and optimization techniques, leading to improved performance in poker.

Contribution

It proposes a unified perspective on deep equilibrium-finding algorithms, introducing new neural modules and optimization methods that outperform existing frameworks.

Findings

Outperforms both PSRO and CFR in Leduc poker

Introduces a novel response oracle with Q, reaching, and baseline values

Develops a method inspired by fictitious play for component optimization

Abstract

Extensive-form games provide a versatile framework for modeling interactions of multiple agents subjected to imperfect observations and stochastic events. In recent years, two paradigms, policy space response oracles (PSRO) and counterfactual regret minimization (CFR), showed that extensive-form games may indeed be solved efficiently. Both of them are capable of leveraging deep neural networks to tackle the scalability issues inherent to extensive-form games and we refer to them as deep equilibrium-finding algorithms. Even though PSRO and CFR share some similarities, they are often regarded as distinct and the answer to the question of which is superior to the other remains ambiguous. Instead of answering this question directly, in this work we propose a unified perspective on deep equilibrium finding that generalizes both PSRO and CFR. Our four main contributions include: i) a novel response oracle (RO) which computes Q values as well as reaching probability values and baseline values; ii) two transform modules -- a pre-transform and a post-transform -- represented by neural networks transforming the outputs of RO to a latent additive space (LAS), and then the LAS to action probabilities for execution; iii) two average oracles -- local average oracle (LAO) and global average oracle (GAO) -- where LAO operates on LAS and GAO is used for evaluation only; and iv) a novel method inspired by fictitious play that optimizes the transform modules and average oracles, and automatically selects the optimal combination of components of the two frameworks. Experiments on Leduc poker game demonstrate that our approach can outperform both frameworks.