Analogies between Software Reverse Engineering and Mechanistic Interpretability — AI Alignment Forum