<ul><li>Designing neural networks typically involves manual trial and error or neural architecture search (NAS) followed by weight training.</li><li>A new approach called SWAT-NN optimizes both the architecture and the weights of a neural network simultaneously.</li><li>This method uses a universal multi-scale autoencoder to embed architectural and parametric information into a continuous latent space.</li><li>Experiments show that SWAT-NN effectively discovers sparse and compact neural networks with strong performance on synthetic regression tasks.</li></ul>

SWAT-NN: Simultaneous Weights and Architecture Training for Neural Networks in a Latent Space

Discover more