No, because that requires the assumption that the model expresses intelligence.