CubeTR: Learning to Solve The Rubiks Cube Using Transformers
This work addresses sparse reward challenges in reinforcement learning for puzzle-solving, offering insights for generalization to higher-dimensional cubes and other sparse reward scenarios, though it is incremental in applying transformers to this domain.
The paper tackles solving the Rubik's Cube using transformers by addressing sparse rewards, achieving solutions from arbitrary states without human priors, with solution lengths close to expert algorithms after move regularization.
Since its first appearance, transformers have been successfully used in wide ranging domains from computer vision to natural language processing. Application of transformers in Reinforcement Learning by reformulating it as a sequence modelling problem was proposed only recently. Compared to other commonly explored reinforcement learning problems, the Rubiks cube poses a unique set of challenges. The Rubiks cube has a single solved state for quintillions of possible configurations which leads to extremely sparse rewards. The proposed model CubeTR attends to longer sequences of actions and addresses the problem of sparse rewards. CubeTR learns how to solve the Rubiks cube from arbitrary starting states without any human prior, and after move regularisation, the lengths of solutions generated by it are expected to be very close to those given by algorithms used by expert human solvers. CubeTR provides insights to the generalisability of learning algorithms to higher dimensional cubes and the applicability of transformers in other relevant sparse reward scenarios.