MediumNLPnlp-075

Softmax with Temperature

Problem

Return softmax(logits / temperature), computed with the max-subtraction trick for numerical stability.

temperature is assumed positive and non-zero. Return an empty list for empty logits.

Hint

Low temperature sharpens the distribution; dividing by it amplifies the logits.

case 1

in: [[1,1],1]

expected: [0.5,0.5]

case 2

in: [[1,2,3],1]

expected: [0.09003057317038046,0.24472847105479764,0.6652409557748218]

case 3

in: [[1,2,3],0.5]

expected: [0.015876239976466765,0.11731042782619838,0.8668133321973349]

case 4

in: [[2,2],0.5]

expected: [0.5,0.5]

case 5

in: [[],1]

expected: []

Starter codePython

import math
def softmax_temperature(logits, temperature):
    # Your code here
    pass

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