A Novel Metric for Measuring the Robustness of Large Language Models in Non-adversarial Scenarios

Open in new window