EasyCalculusca-278

Softplus Derivative Vector

Problem

The softplus function f(z) = log(1 + exp(z)) has derivative sigmoid(z) = 1 / (1 + exp(-z)).

Return the derivative at each of the given values. Use a numerically stable sigmoid.

Hint

Softplus differentiates to the logistic function; branch to avoid overflow.

case 1

in: [[0,1,-1]]

expected: [0.5,0.7310585786300049,0.2689414213699951]

case 2

in: [[10,-10]]

expected: [0.9999546021312976,0.000045397868702434395]

case 3

in: [[0.5,-0.5,2]]

expected: [0.6224593312018546,0.37754066879814546,0.8807970779778823]

Starter codePython

def softplus_derivative(values):
    # Your code here
    pass

3 test cases · run your code against them in the editor