Bases: torch.nn.modules.module.Module
torch.nn.modules.module.Module
K-winners-take-all activation function.
k, the number of active (winner) neurons within an output layer. Default: 50
Methods
forward(x)
forward
The forward pass of kWTA.
Attributes
An input vector.
The output vector y = kwta(x) with exactly k active neurons.
y = kwta(x)
k