🔎 CRUXEval Sample Explorer 🔎

CRUXEval is a benchmark complementary to HumanEval and MBPP measuring code reasoning, understanding, and execution capabilities!


   
def f(text):
    text = list(text)
    for i in range(len(text)):
        if i % 2 == 1:
            text[i] = text[i].swapcase()
    return ''.join(text)
assert f('Hey DUdE THis $nd^ &*&this@#') == 'HEy Dude tHIs $Nd^ &*&tHiS@#'
CRUXEval-I
assert f(??) == 'HEy Dude tHIs $Nd^ &*&tHiS@#'

CRUXEval-O
assert f('Hey DUdE THis $nd^ &*&this@#') == ??