#Exploration: A Study of Count-Based Exploration for Deep Reinforcement Learning

TL;DR

This paper demonstrates that a simple count-based exploration method using hash codes can achieve near state-of-the-art results in high-dimensional and continuous deep reinforcement learning tasks, challenging the belief that such methods are limited to small state spaces.

Contribution

The authors introduce a straightforward extension of count-based exploration using hash codes, which performs well on complex high-dimensional and continuous RL benchmarks, and analyze key properties of effective hash functions.

Findings

Hash-based exploration achieves near state-of-the-art results on deep RL benchmarks.

Simple hash functions can effectively guide exploration in high-dimensional spaces.

Domain-dependent learned hash codes can further enhance exploration performance.

Abstract

Count-based exploration algorithms are known to perform near-optimally when used in conjunction with tabular reinforcement learning (RL) methods for solving small discrete Markov decision processes (MDPs). It is generally thought that count-based methods cannot be applied in high-dimensional state spaces, since most states will only occur once. Recent deep RL exploration strategies are able to deal with high-dimensional continuous state spaces through complex heuristics, often relying on optimism in the face of uncertainty or intrinsic motivation. In this work, we describe a surprising finding: a simple generalization of the classic count-based approach can reach near state-of-the-art performance on various high-dimensional and/or continuous deep RL benchmarks. States are mapped to hash codes, which allows to count their occurrences with a hash table. These counts are then used to compute a reward bonus according to the classic count-based exploration theory. We find that simple hash functions can achieve surprisingly good results on many challenging tasks. Furthermore, we show that a domain-dependent learned hash code may further improve these results. Detailed analysis reveals important aspects of a good hash function: 1) having appropriate granularity and 2) encoding information relevant to solving the MDP. This exploration strategy achieves near state-of-the-art performance on both continuous control tasks and Atari 2600 games, hence providing a simple yet powerful baseline for solving MDPs that require considerable exploration.