The Hidden Risks of Large Reasoning Models: A Safety Assessment of R1

Zhou, Kaiwen, Liu, Chengzhi, Zhao, Xuandong · arXiv · 2025