Neel Nanda comments on How To Become A Mechanistic Interpretability Researcher