Exploring the Robustness of Model-Graded Evaluations and Automated Interpretability

Open in new window