CHATS-Lab/MT-AgentRisk download history
CHATS-Lab/MT-AgentRisk is an other dataset on the Hugging Face Hub. In the last 30 days it was downloaded 328 times (49 in the last 7 days), and 2,437 times in total. It ranks #45,768 among datasets by monthly downloads.
MT-AgentRisk Multi-Turn Agent Safety Benchmark 📄Paper 🌐Website 💻GitHub Overview MT-AgentRisk is the first benchmark for evaluating multi-turn safety risks in tool-using LLM agents. As agents become increasingly capable, their safety lags behind — creating a widening