Skip to content

Translation in progress Some entries still include text in the original language.

Oct 12 · Monday

All papers

Found 3 papers
Filter3

Impact

Unauthorized actions15Data and privacy leaks3Output manipulation5Oversight failure3

System type

AI Agent3

Attack surface / entry point

User input4Code repositories and dependencies3

Layer

Application layer3
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 265papers have no tags yet and are not included in filtered results.
Impact：Oversight failureSystem type：AI AgentAttack surface / entry point：Code repositories and dependenciesClear all

Another 265papers do not have category tags yet and are not included in filtered results.

  2. DefensesarXiv 2609.35659

    Tracekit：面向自主编码 Agent 的防篡改意图-推理-动作审计系统

    提出 Tracekit，审计编码 Agent 的意图、推理与动作并检测账本篡改

    arXiv
    2609.35659
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    Claude Code (claude -p, version 2.1), 独立评审（同一 CLI，无工具）
    AbstractTracekit 把意图、自述和动作写入可锚定的哈希链。

    Tracekit 是无第三方依赖的编码智能体审计层：把人类请求、模型自述和已执行动作并排写入可外部锚定的哈希链，执行前用正则门，事后用规则和独立模型互查。

    Paper analysisRead analysis
  3. AttacksarXiv 2605.12673

    UC Berkeley proposes BenchJack, which automatically audits 10 agent benchmarks and finds 219 reward hacking defects.

    提出 BENCHJACK，审计 Agent 基准中的奖励作弊缺陷

    arXiv
    2605.12673
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Abstract论文系统揭示了智能体基准的奖励作弊隐患，提出自动化红队审计与迭代修补系统 BENCHJACK。

    论文系统研究了主流 AI 智能体基准评测中的奖励作弊问题，提出了首个针对基准缺陷的自动化红队审计与生成对抗式迭代修补系统 BENCHJACK。

    Paper analysisRead analysis
You have reached the end