Adaptive Computation Time for Recurrent Neural NetworksThis paper introduces Adaptive Computation Time (ACT), an algorithm that allows recurrent neural networks to learn how many computational steps to take between receiving an input and emitting an output. ACT requires minimal changes to the network architecture, is deterministic and differentiable, and does not add any noise to the parameter gradients. Experimental results are provided for four synt
Pixel Recurrent Neural Networks
Google Sites: Sign-in
