AI.info

Shauli Ravfogel

Explore Shauli Ravfogel on AI.info.

• Making LLMs Say What They Think: Measuring and Improving CoT-Interpretability Alignment
• Emergence of Linear Truth Encodings in Language Models