Solving the Rubik's Cube Without Human Knowledge
This addresses the challenge of sparse rewards in combinatorial optimization for AI agents, representing a significant advance over prior methods that relied on human domain knowledge.
The paper tackles the problem of solving the Rubik's Cube without human knowledge by introducing Autodidactic Iteration, a novel reinforcement learning algorithm that achieves 100% success on randomly scrambled cubes with a median solve length of 30 moves.
A generally intelligent agent must be able to teach itself how to solve problems in complex domains with minimal human supervision. Recently, deep reinforcement learning algorithms combined with self-play have achieved superhuman proficiency in Go, Chess, and Shogi without human data or domain knowledge. In these environments, a reward is always received at the end of the game, however, for many combinatorial optimization environments, rewards are sparse and episodes are not guaranteed to terminate. We introduce Autodidactic Iteration: a novel reinforcement learning algorithm that is able to teach itself how to solve the Rubik's Cube with no human assistance. Our algorithm is able to solve 100% of randomly scrambled cubes while achieving a median solve length of 30 moves -- less than or equal to solvers that employ human domain knowledge.