风险评测 · 评测基准

Quantifying Frontier LLM Capabilities for Container Sandbox Escape

NUWA 研究解读 · 独立复核待完成

首次发表: 2026-03-23来源核验: 2026-10-25

原始来源 ↗引用本记录 ↓网络安全 →

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：该基准探测特定弱点条件下的隔离失败。

研究问题

受控条件下，智能体能否逃出刻意弱化的容器沙箱？

研究设计与假设

嵌套隔离设计让有Shell权限的对抗智能体测试内层弱点，外层限制实验影响。

SandboxEscapeBench — 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

更新结果报告Opus 4.5在56%样本上成功。

Mythos Preview每个样本至少成功一次，不等于单次运行100%成功（§6.1）。

NUWA 解读

NUWA 解读：该基准探测特定弱点条件下的隔离失败。

局限与相反证据

最少监测、人为弱点与专家估难不代表有防护生产环境的失陷概率（§7）。

研究与决策意义

模型建议：区分原始与更新结果，并保留重复次数语义。

精确来源位置

Quantifying Frontier LLM Capabilities for Container Sandbox Escape · AbstractSelected original resource https://arxiv.org/html/2603.02277; sha256:078f48e266792c2b63ee79c3425750429f323b2cf104dacbfa452ce6774da4e7; archived source capture; exact capture time not independently established
Quantifying Frontier LLM Capabilities for Container Sandbox Escape · §3 MethodologySelected original resource https://arxiv.org/html/2603.02277; sha256:078f48e266792c2b63ee79c3425750429f323b2cf104dacbfa452ce6774da4e7; archived source capture; exact capture time not independently established
Quantifying Frontier LLM Capabilities for Container Sandbox Escape · §6.1 Primary ResultsSelected original resource https://arxiv.org/html/2603.02277; sha256:078f48e266792c2b63ee79c3425750429f323b2cf104dacbfa452ce6774da4e7; archived source capture; exact capture time not independently established
Quantifying Frontier LLM Capabilities for Container Sandbox Escape · §7 LimitationsSelected original resource https://arxiv.org/html/2603.02277; sha256:078f48e266792c2b63ee79c3425750429f323b2cf104dacbfa452ce6774da4e7; 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

研究机构牛津大学英国 AI 安全研究所

作者 (11)

Rahul Marchand · Art O Cathain · Jerome Wynne · Philippos Maximos Giavridis · Stuart Jennings · Freddy Tuxworth · Tolga H. Dur · Sam Deverett · John Wilkinson · Jason Gwartz · Harry Coppock

评测协议

SandboxEscapeBench →

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{sandboxescapebench,
  title = {Quantifying Frontier LLM Capabilities for Container Sandbox Escape},
  url = {https://www.aisi.gov.uk/research/quantifying-frontier-llm-capabilities-for-container-sandbox-escape},
  author = {Rahul Marchand and Art O Cathain and Jerome Wynne and Philippos Maximos Giavridis and Stuart Jennings and Freddy Tuxworth and Tolga H. Dur and Sam Deverett and John Wilkinson and Jason Gwartz and Harry Coppock},
  year = {2026},
  date = {2026-03-23},
  version = {AISI research page checked 2026-10-25},
  note = {AISI research page checked 2026-10-25}
}

NUWA 记录 · BibTeX

@misc{sandboxescapebench-nuwa,
  title = {Quantifying Frontier LLM Capabilities for Container Sandbox Escape — NUWA research record},
  url = {https://whitzard.tech/nuwa/whitzard-index/work/sandboxescapebench/},
  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--sandboxescapebench.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--sandboxescapebench.json.}
}

证据快照 2026-10-29.1 · 索引更新 · RSS

参考文献

[1] Rahul Marchand, Art O Cathain 等. Quantifying Frontier LLM Capabilities for Container Sandbox Escape. AISI research page checked 2026-10-25 ; Retained existing source-specific note sections; this pass audits duplicate/boilerplate structure and source availability, not complete section-level verification.
[2] Rahul Marchand, Art O Cathain 等. Quantifying Frontier LLM Capabilities for Container Sandbox Escape ; Author affiliations on the original paper title page; Abstract; §3 Methodology