Comment by porridgeraisin

It does not, if it gets the answer (or any information about them, even % of qns solved) and is able to adjust itself in response, then that is considered training on the test set and is wrong.