Divergence Detect In Off-Policy FA

Problem

Detect divergence by checking whether any value in the history exceeds the threshold in absolute value. Return True on the first violation.

Hint

The check is strict: values exactly at the threshold are still considered stable.

case 1

in: [[1,2,3],10]

expected: false

case 2

in: [[1,-50,3],10]

expected: true

case 3

in: [[],1]

expected: false

case 4

in: [[100],100]

expected: false

Starter codePython

def divergence_detected(history, threshold):
    # Your code here
    pass

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