NALU implementation and experiments: 
A TensorFlow and Keras implementation of neural arithmetic logic unit from arXiv:1808.00508. The neural arithmetic logic unit is a newly developed logic unit that attempts to improve a neural network’s competence to handle numerical data.
   Mixture of Softmaxes implementation and experiments: 
A Keras implementation Mixture of Softmaxes. MoS layer is a way training an ensemble method without the having to train the full NN. This layer is described in Breaking the Softmax Bottleneck.

Contribution chart
acyrl's Github Chart