AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents

Open in new window