Solving Sokoban with Forward-Backward Reinforcement Learning.

Yaron Shoham, Gal Elidan

Browse the full SoCS paper archive.