Feb 26, 2026 Understanding Language Models 1: Mechanistic Interpretability Meets Causal Representation Learning