Activation Explorer

Visualize how neural networks introduce non-linearity. Manipulate the input signal z to see how different functions transform it.

Active Function

Sigmoid

Tanh

ReLU

Leaky ReLU

Step

Properties & Formula

Range: (0, 1) Deriv at z=0: 0.25 Behavior: Saturates at extremes

Controls

Input Signal (z) 0.00

Show Derivative f'(z)

Compare All Functions