Logan Riggs comments on ‘Fundamental’ vs ‘applied’ mechanistic interpretability research