Learning to Repeat: Fine Grained Action Repetition for Deep Reinforcement Learning

TL;DR

This paper introduces FiGAR, a framework that allows reinforcement learning agents to decide not only which action to take but also how long to repeat it, enhancing performance across various algorithms and domains.

Contribution

The paper presents FiGAR, a novel approach for temporal abstraction in deep reinforcement learning, improving existing algorithms by enabling action repetition decisions.

Findings

Performance improvements on Atari 2600 with A2C

Enhanced results in Mujoco with TRPO

Better outcomes in TORCS with DDPG

Abstract

Reinforcement Learning algorithms can learn complex behavioral patterns for sequential decision making tasks wherein an agent interacts with an environment and acquires feedback in the form of rewards sampled from it. Traditionally, such algorithms make decisions, i.e., select actions to execute, at every single time step of the agent-environment interactions. In this paper, we propose a novel framework, Fine Grained Action Repetition (FiGAR), which enables the agent to decide the action as well as the time scale of repeating it. FiGAR can be used for improving any Deep Reinforcement Learning algorithm which maintains an explicit policy estimate by enabling temporal abstractions in the action space. We empirically demonstrate the efficacy of our framework by showing performance improvements on top of three policy search algorithms in different domains: Asynchronous Advantage Actor Critic in the Atari 2600 domain, Trust Region Policy Optimization in Mujoco domain and Deep Deterministic Policy Gradients in the TORCS car racing domain.