JuniorTheoryCommon

What does this program print?

Python

import numpy as np

x = np.array([1.0, 2.0])
W = np.array([[1.0, -2.0], [1.0, 0.5]])
b = np.array([0.5, 0.5])
h = np.maximum(0, x @ W + b)
print(h[0], h[1])

3.5 0.0

Expecting 3.5 -0.5: ReLU replaces every negative sum with zero, so the second neuron's -0.5 becomes 0.0.
Expecting 0.0 2.5: in x @ W each column of W holds one neuron's weights, so the first neuron is 1 * 1 + 2 * 1 + 0.5 = 3.5.

What would the program print if ReLU were replaced by the sigmoid?
Why does stacking layers without any activation give nothing a single linear layer lacks?

Back to questions Open topic in roadmap →Answer it yourself — Quiz

Back to questions

JuniorTheoryCommon

What does this program print?

Python

import numpy as np

x = np.array([1.0, 2.0])
W = np.array([[1.0, -2.0], [1.0, 0.5]])
b = np.array([0.5, 0.5])
h = np.maximum(0, x @ W + b)
print(h[0], h[1])

3.5 0.0

Expecting 3.5 -0.5: ReLU replaces every negative sum with zero, so the second neuron's -0.5 becomes 0.0.
Expecting 0.0 2.5: in x @ W each column of W holds one neuron's weights, so the first neuron is 1 * 1 + 2 * 1 + 0.5 = 3.5.

What would the program print if ReLU were replaced by the sigmoid?
Why does stacking layers without any activation give nothing a single linear layer lacks?

Back to questions Open topic in roadmap →Answer it yourself — Quiz