EIS VI: Critiques of Mechanistic Interpretability Work in AI Safety — AI Alignment Forum