I am interested in Alignment, Mechanistic Interpretability, Agents, and the theory of how neural networks work.