Why would that be relevant? I mean tests like that mostly reward memorising knowledge, something that is easy for a computer and hard for a human. We use it as

Why would that be relevant? I mean tests like that mostly reward memorising knowledge, something that is easy for a computer and hard for a human. We use it as a heuristic for having studied the knowledge in depth since it scales and is relatively fair, not because it's actually a good way to evaluate competency. Of we could do completely fair (i.e. comparable between candidates), comprehensive 1-1 oral exams of every student without cost, we probably would do that a lot more. This heuristic does not work for LLMs, so imho the use of this benchmark seems extremely limited.