Infinitely wide limits for deep Stable neural networks: sub-linear, linear and super-linear activation functions

Open in new window