Python Idle
Sigmoid squashes. ReLU clips. Tanh centers. Without activation functions, a neural network is just a fancy linear regression — no matter how many layers you stack, the result is still a straight line. Activations inject the nonlinearity that lets networks learn curves, edges, and abstractions. Today you'll implement the three most important ones and visualize exactly how they transform a signal.
~15 minsandbox lab
Loading Python runtime...
Goals: 6 tests
sigmoid(0) should be approximately 0.5
relu(-1) should be 0
relu(1) should be 1
tanh(0) should be approximately 0
all activation functions should be callable
should create a matplotlib plot
Python loading...