IGN : Implicit Generative Networks

TL;DR

This paper introduces a novel distributional reinforcement learning method called IGN, which combines GANs with quantile regression to improve the modeling of return distributions, demonstrating superior performance on Atari games.

Contribution

The paper presents a new distributional RL algorithm that integrates GANs with quantile regression, achieving state-of-the-art results in Atari game benchmarks.

Findings

Improved performance on 57 Atari games

State-of-the-art risk-sensitive policy training

Effective modeling of return distributions

Abstract

In this work, we build recent advances in distributional reinforcement learning to give a state-of-art distributional variant of the model based on the IQN. We achieve this by using the GAN model's generator and discriminator function with the quantile regression to approximate the full quantile value for the state-action return distribution. We demonstrate improved performance on our baseline dataset - 57 Atari 2600 games in the ALE. Also, we use our algorithm to show the state-of-art training performance of risk-sensitive policies in Atari games with the policy optimization and evaluation.