Softmax

• 
• Why is this?
  • We want the sum of the outputs to be equal to 1, so we express it as a fraction of the total sum.
  • Additionally, we want the value of the input z to be positive regardless of its value, so we use $e^z$.
• The result is in the same format as a One-hot vector.
  • A one-hot vector can be seen as the output of softmax, effectively indicating that one probability is 100%.