On the Validity of Traditional Vulnerability Scoring Systems for Adversarial Attacks against LLMs

Open in new window