Attention and memory in deep learning

68