InnerThoughts: Disentangling Representations and Predictions in Large Language Models

Open in new window