Reward-based learning of sensorimotor synchronization in recurrent neural networks Posters