harmful-content-detection

# 超越静态基准：基于角色模拟合成有害内容以实现鲁棒性评估 Source: [https://arxiv.org/html/2604.17020](https://arxiv.org/html/2604.17020) Huije Lee Jisu Shin Hoyun Song Changgeon Ko Jong C\. Park Korea Advanced Institute of Science and Technology \(KAIST\) \{huijelee,jisu\.shin,hysong,pencaty,jongpark\}@kaist\.ac\.kr ###### Abstract 面向有害内容检测的静态基准在可扩展性与多样性方面存在局限，且可能受...