Gravar-mail: Correction: Temporal-Difference Reinforcement Learning with Distributed Representations