No, what is rewarded is "the code has been shown to conform to the given objective function" and especially "that objective function is a good representation of what we are trying to accomplish with this code".