🔎 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(text1, text2):
    nums = []
    for i in range(len(text2)):
        nums.append(text1.count(text2[i]))
    return sum(nums)
assert f('jivespdcxc', 'sx') == 2
CRUXEval-I 
assert f(??) == 2
 

            		
          

          

            
              CRUXEval-O
              
assert f('jivespdcxc', 'sx') == ??

              

              

            		
          

        

      


      
      

      

        
        
        
        

          564 / 800
        

      


      
      

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