Datasets:

openai_humaneval

like 184

" def is_multiply_prime(a): """Write a function that returns true if the given number is the multiplication of 3 prime numbers and false otherwise. Knowing that (a) is less then 100. Example: is_multiply_prime(30) == True 30 = 2 * 3 * 5 """ "

This function specifically says 'Knowing that (a) is less then 100.'. But then if you check to see the tests they run, they do many numbers that are over 100 and expect correct results for those.

Seems like there are a bunch of badly written prompts/tests in this dataset.