Adaptive Hybrid Activation Function

Introduced by Bodyanskiy et al. in Adaptive hybrid activation function for deep neural networks

Trainable activation function as a sigmoid-based generalization of ReLU, Swish and SiLU.

Source: Adaptive hybrid activation function for deep neural networks

Paper	Code	Results	Date	Stars

Task	Papers	Share
Activation Function Synthesis	2	33.33%
Image Classification	2	33.33%
Learning Theory	2	33.33%

This feature is experimental; we are continuously improving our matching algorithm.