🔎 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(text):
    rtext = list(text)
    for i in range(1, len(rtext) - 1):
        rtext.insert(i + 1, '|')
    return ''.join(rtext)
assert f('pxcznyf') == 'px|||||cznyf'
CRUXEval-I 
assert f(??) == 'px|||||cznyf'
 

            		
          

          

            
              CRUXEval-O
              
assert f('pxcznyf') == ??

              

              

            		
          

        

      


      
      

      

        
        
        
        

          308 / 800
        

      


      
      

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