🔎 CRUXEval Sample Explorer 🔎

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

Homepage Paper Code HF Dataset Leaderboard

    
def f(numbers, index):
    for n in numbers[index:]:
        numbers.insert(index, n)
        index += 1
    return numbers[:index]
assert f([-2, 4, -4], 0) == [-2, 4, -4]
CRUXEval-I 
assert f(??) == [-2, 4, -4]
 

            		
          

          

            
              CRUXEval-O
              
assert f([-2, 4, -4], 0) == ??

              

              

            		
          

        

      


      
      

      

        
        
        
        

          151 / 800
        

      


      
      

        The site is inspired by and based on
        Minerva sample explorer and
        Llemma. Thanks!