Language Models 2 The Polysemy of Mechanistic Interpretability Sep 6, 2026 Transformer and Hallucinations in Simple Language Jul 5, 2024