I thought that at least some ZKPs rely on responses to one or a set of specific challenge(s) randomly chosen by the interrogator. So, as long as the challenge s

I thought that at least some ZKPs rely on responses to one or a set of specific challenge(s) randomly chosen by the interrogator. So, as long as the challenge space is big enough, it seems unlikely that you can easily collect enough responses to satisfy such a random (set of) challenge(s) with any significant probability. Please correct me, if I got that wrong.