Mechanistic Interpretability of Emotion Inference in Large Language Models

Open in new window