[retinaros]

we dont know what OAI does to find the best answer when reasoning but I am pretty sure that having variations of a same model is part of it.