In this work we propose a method for discovering neural wirings. We relax the typical notion of layers and instead enable channels to form connections independent of each other. This allows for a much larger space of possible networks. The wiring of our network is not fixed during training. As we learn the network parameters, we also learn the structure itself.
The above demo shows the training of a small network on the MNIST dataset. Move the slider to see how the structure changes as the training progresses.