风险评测 · 评测基准

SHADE-Arena: Evaluating sabotage and monitoring in LLM agents

NUWA 研究解读 · 独立复核待完成

首次发表: 2025-06-16来源核验: 2026-10-25

风险监测

原始来源 ↗引用本记录 ↓欺骗 →

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：监测质量是运行阈值权衡，不是单一检出百分比。

研究问题

监测器能否识别合法工具任务中隐藏的蓄意破坏？

研究设计与假设

封闭实验环境给智能体合法任务与隐蔽副任务，另一监测模型判断行为。

SHADE-Arena — Figure 3 — Figure 3 · 原文 Figure 3。请结合下方的协议范围与局限阅读。原图与来源 ↗

来源发现

最佳受测监测器Gemini 2.5 Pro要达到80%检出，需要超过20%误报。

提示调整与轨迹微调有时可改善监测（结果）。

NUWA 解读

NUWA 解读：监测质量是运行阈值权衡，不是单一检出百分比。

局限与相反证据

这是模拟中诱导的副任务；作者认为所需误报率不适合实际部署。

研究与决策意义

模型建议：保留真阳性、假阳性工作点，并区分诱导与自发破坏。

精确来源位置

SHADE-Arena: Evaluating sabotage and monitoring in LLM agents · SHADE-ArenaSelected original resource https://www.anthropic.com/research/shade-arena-sabotage-monitoring; sha256:0a989a6b3149554677b8b01058a0d2b97da95a26b0063898bfa438824d8ed08d; archived source capture; exact capture time not independently established
SHADE-Arena: Evaluating sabotage and monitoring in LLM agents · ResultsSelected original resource https://www.anthropic.com/research/shade-arena-sabotage-monitoring; sha256:0a989a6b3149554677b8b01058a0d2b97da95a26b0063898bfa438824d8ed08d; archived source capture; exact capture time not independently established
SHADE-Arena: Evaluating sabotage and monitoring in LLM agents · ConclusionsSelected original resource https://www.anthropic.com/research/shade-arena-sabotage-monitoring; sha256:0a989a6b3149554677b8b01058a0d2b97da95a26b0063898bfa438824d8ed08d; 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

发布机构 Anthropic

评测协议

SHADE-Arena →

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{shade-arena,
  title = {SHADE-Arena: Evaluating sabotage and monitoring in LLM agents},
  url = {https://www.anthropic.com/research/shade-arena-sabotage-monitoring},
  year = {2025},
  date = {2025-06-16},
  version = {Anthropic research article 2025-06-16},
  note = {Anthropic research article 2025-06-16}
}

NUWA 记录 · BibTeX

@misc{shade-arena-nuwa,
  title = {SHADE-Arena: Evaluating sabotage and monitoring in LLM agents — NUWA research record},
  url = {https://whitzard.tech/nuwa/whitzard-index/work/shade-arena/},
  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--shade-arena.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--shade-arena.json.}
}

证据快照 2026-10-29.1 · 索引更新 · RSS

参考文献

[1] SHADE-Arena: Evaluating sabotage and monitoring in LLM agents. 2025-06-16 ; Official publication masthead and article byline; SHADE-Arena; Results
[2] SHADE-Arena: Evaluating sabotage and monitoring in LLM agents. 2025-06-16