Guardrail evaluation¶ Systematic evaluation and benchmarking of safety mechanisms and guardrails designed to prevent harmful or misinformed outputs from language models. Related topics¶ LLM Safety and Adversarial Robustness Fact-checking and corrections Large language model evaluation