List of Paper for AI Agent Safety Benchmark
-
AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents
Paper • 2410.09024 • Published • 1 -
Agent Security Bench (ASB): Formalizing and Benchmarking Attacks and Defenses in LLM-based Agents
Paper • 2410.02644 • Published -
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Paper • 2402.04249 • Published • 6
