Scalable Online Planning via Reinforcement Learning Fine-Tuning

TL;DR

This paper introduces a reinforcement learning-based online fine-tuning method for policy neural networks that replaces traditional tabular search, achieving superior performance in complex, stochastic, and partially observable environments like Hanabi and Ms. Pacman.

Contribution

It presents a scalable online planning approach using reinforcement learning to fine-tune policies, outperforming traditional tabular search methods in benchmark tasks.

Findings

Achieved new state-of-the-art in self-play Hanabi.

Outperformed tabular search in Atari Ms. Pacman.

Demonstrated generality across different game environments.

Abstract

Lookahead search has been a critical component of recent AI successes, such as in the games of chess, go, and poker. However, the search methods used in these games, and in many other settings, are tabular. Tabular search methods do not scale well with the size of the search space, and this problem is exacerbated by stochasticity and partial observability. In this work we replace tabular search with online model-based fine-tuning of a policy neural network via reinforcement learning, and show that this approach outperforms state-of-the-art search algorithms in benchmark settings. In particular, we use our search algorithm to achieve a new state-of-the-art result in self-play Hanabi, and show the generality of our algorithm by also showing that it outperforms tabular search in the Atari game Ms. Pacman.