A Novel Automated Curriculum Strategy to Solve Hard Sokoban Planning Instances

TL;DR

This paper introduces an automated curriculum learning strategy guided by difficulty quantum momentum, significantly enhancing reinforcement learning performance on complex Sokoban planning problems by enabling the solution of previously intractable instances.

Contribution

The paper presents a novel automated curriculum approach that dynamically selects training instances based on difficulty, improving RL performance on complex combinatorial planning tasks like Sokoban.

Findings

RL agent solves Sokoban instances previously unsolvable by state-of-the-art methods.

Automated curriculum reduces difficulty gap, enabling smoother learning.

Coupling curriculum with curiosity-driven search further boosts performance.

Abstract

In recent years, we have witnessed tremendous progress in deep reinforcement learning (RL) for tasks such as Go, Chess, video games, and robot control. Nevertheless, other combinatorial domains, such as AI planning, still pose considerable challenges for RL approaches. The key difficulty in those domains is that a positive reward signal becomes {\em exponentially rare} as the minimal solution length increases. So, an RL approach loses its training signal. There has been promising recent progress by using a curriculum-driven learning approach that is designed to solve a single hard instance. We present a novel {\em automated} curriculum approach that dynamically selects from a pool of unlabeled training instances of varying task complexity guided by our {\em difficulty quantum momentum} strategy. We show how the smoothness of the task hardness impacts the final learning results. In particular, as the size of the instance pool increases, the ``hardness gap'' decreases, which facilitates a smoother automated curriculum based learning process. Our automated curriculum approach dramatically improves upon the previous approaches. We show our results on Sokoban, which is a traditional PSPACE-complete planning problem and presents a great challenge even for specialized solvers. Our RL agent can solve hard instances that are far out of reach for any previous state-of-the-art Sokoban solver. In particular, our approach can uncover plans that require hundreds of steps, while the best previous search methods would take many years of computing time to solve such instances. In addition, we show that we can further boost the RL performance with an intricate coupling of our automated curriculum approach with a curiosity-driven search strategy and a graph neural net representation.