CRUXEval Sample Explorer

model: temp:

def f(book):
    a = book.rsplit(':', 1)
    if a[0].split(' ')[-1] == a[1].split(' ')[0]:
        return f(' '.join(a[0].split(' ')[:-1]) + ' ' + a[1])
    return book
assert f("udhv zcvi nhtnfyd :erwuyawa pun") == 'udhv zcvi nhtnfyd :erwuyawa pun'

CRUXEval-I assert f(??) == 'udhv zcvi nhtnfyd :erwuyawa pun'

CRUXEval-O assert f("udhv zcvi nhtnfyd :erwuyawa pun") == ??

🔎 CRUXEval Sample Explorer 🔎

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