Razied comments on Using GPT-N to Solve Interpretability of Neural Networks: A Research Agenda