风险评测 · 评测基准

Agent-SafetyBench: Evaluating the Safety of LLM Agents

NUWA 研究解读 · 独立复核待完成

首次发表: 2024-12-19来源核验: 2026-10-25

原始来源 ↗引用本记录 ↓

NUWA Research Note

研究问题研究设计与假设来源发现 NUWA 解读局限与相反证据研究与决策意义机构与角色原始资源

模型辅助整理 · 选读原始全文或官方文章中的相关章节；尚未完成逐章节审计。

研究速读

NUWA 解读：稳健性与风险意识是安全交互的不同要求。

研究问题

智能体与环境、工具交互时有哪些安全故障？

研究设计与假设

349环境、2,000案例覆盖八风险类及十故障模式，细化、去重并统一既有案例。

Agent-SafetyBench — Figure 1 — Figure 1 · 原文 Figure 1。请结合下方的协议范围与局限阅读。原图与来源 ↗

来源发现

16个受测智能体总安全分均未超60%。

防御提示收益有限，部分弱模型没有改善（§4.5）。

NUWA 解读

NUWA 解读：稳健性与风险意识是安全交互的不同要求。

局限与相反证据

构造场景及特定判分不等于部署事故率，高能力也不保证安全。

研究与决策意义

模型建议：查看故障模式与帮助性明细，不只看总分。

精确来源位置

Agent-SafetyBench : Evaluating the Safety of LLM Agents · AbstractSelected original resource https://arxiv.org/html/2412.14470; sha256:d75e5fb230a89a9ffa0484fbfe694ac609b7b5de3aa59204cadf73b473737338; archived source capture; exact capture time not independently established
Agent-SafetyBench : Evaluating the Safety of LLM Agents · §3.2.2 Refine Existing DatasetsSelected original resource https://arxiv.org/html/2412.14470; sha256:d75e5fb230a89a9ffa0484fbfe694ac609b7b5de3aa59204cadf73b473737338; archived source capture; exact capture time not independently established
Agent-SafetyBench : Evaluating the Safety of LLM Agents · §4.2 Main ResultsSelected original resource https://arxiv.org/html/2412.14470; sha256:d75e5fb230a89a9ffa0484fbfe694ac609b7b5de3aa59204cadf73b473737338; archived source capture; exact capture time not independently established
Agent-SafetyBench : Evaluating the Safety of LLM Agents · §4.5 Discussion on DefenseSelected original resource https://arxiv.org/html/2412.14470; sha256:d75e5fb230a89a9ffa0484fbfe694ac609b7b5de3aa59204cadf73b473737338; archived source capture; exact capture time not independently established

来源核查、数据提取、编辑批准与实验复现分别记录。方法说明 →

机构与角色

研究机构清华大学

作者 (7)

Zhexin Zhang · Shiyao Cui · Yida Lu · Jingzhuo Zhou · Junxiao Yang · Hongning Wang · Minlie Huang

评测协议

Agent-SafetyBench →

引用与复用

引用研究发现时请引用原始出版物；引用 NUWA 的整理与解释时，单独引用本站记录及对应快照。

BibTeX ↓RIS ↓记录与来源元数据 · JSON ↓阅读覆盖 · JSON 提交更正

原始出版物 · BibTeX

@misc{agent-safetybench,
  title = {Agent-SafetyBench: Evaluating the Safety of LLM Agents},
  url = {https://arxiv.org/abs/2412.14470},
  author = {Zhexin Zhang and Shiyao Cui and Yida Lu and Jingzhuo Zhou and Junxiao Yang and Hongning Wang and Minlie Huang},
  year = {2024},
  date = {2024-12-19},
  version = {arXiv:2412.14470v2},
  note = {arXiv:2412.14470v2}
}

NUWA 记录 · BibTeX

@misc{agent-safetybench-nuwa,
  title = {Agent-SafetyBench: Evaluating the Safety of LLM Agents — NUWA research record},
  url = {https://whitzard.tech/nuwa/whitzard-index/work/agent-safetybench/},
  version = {2026-10-29.1},
  organization = {NUWA Frontier AI Safety Lab},
  howpublished = {Retained note edition: https://whitzard.tech/nuwa/whitzard-index/data/record-editions/2026-10-29.1--agent-safetybench.json},
  note = {NUWA research note. Source-reading scope and independent review status are stated on the record. Snapshot 2026-10-29.1. Retained note edition: https://whitzard.tech/nuwa/whitzard-index/data/record-editions/2026-10-29.1--agent-safetybench.json.}
}

证据快照 2026-10-29.1 · 索引更新 · RSS

参考文献

[1] Zhexin Zhang 1 1 1 Equal contribution., Shiyao Cui 1 1 1 Equal contribution. 等. Agent-SafetyBench: Evaluating the Safety of LLM Agents. 2024-12-19. v2 [arXiv]; Author affiliations on the original paper title page; Abstract; §3.2.2 Refine Existing Datasets