Skip to content

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

Oct 12 · Monday

All papers

Found 6 papers
Filter4

Attack type

Jailbreaks3

Impact

Harmful content6Unauthorized actions37Data and privacy leaks7Output manipulation9Availability and reliability6Oversight failure25

System type

Models and APIs77AI Agent6

Attack surface / entry point

User input3

Layer

Application layer4

System component

Guardrails and judges6Monitors2Vision4Multimodal4
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 261papers have no tags yet and are not included in filtered results.
Impact：Harmful contentSystem type：AI AgentSystem component：MonitorsSystem component：MultimodalClear all

Another 261papers do not have category tags yet and are not included in filtered results.

  1. Evaluation and benchmarksarXiv 2610.07089

    研究者提出统一滥用监控基准，覆盖分解攻击与提示注入约 6200 条轨迹

    提出危害窗口与区间指标，统一评测分解攻击和提示注入下的智能体滥用监视

    arXiv
    2610.07089
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Gemma 4 26B-A4B, Qwen3.6 27B, Qwen3.5 27B and others12

    Abstract统一智能体滥用监视框架，动作视角兼顾两类威胁但危害定位仍存瓶颈。

    Paper analysisRead analysis
  2. Evaluation and benchmarksarXiv 2610.03938

    Paper finds multimodal models' refusal failure rate rises by up to 68.7% after tool use is enabled

    评测 ReAct 工具调用对多模态模型拒答有害请求的削弱

    arXiv
    2610.03938
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Gemini-2.5-Pro, Gemini-3.1-Pro-Preview, Claude Opus 4.6 and others13
    Abstract论文揭示了工具调用范式削弱 MLLM 拒答能力的现象，分析了稀释与偏移机制并验证了重注缓解方案。

    论文系统评估了智能体工具调用范式对多模态大语言模型拒答能力的影响，指出了工具增强视觉推理背后的安全隐患。核心探讨的问题在于：当 MLLM 从单次推理转向通过 ReAct 循环调用外部工具（打标、缩放、OCR、代码执行）时，工具使用机制本身是否会降低模型拒答有害请求的概率。

    Paper analysisRead analysis
  3. Risky behaviorarXiv 2609.38971

    研究测量并预测具身 VLM 规划器中任务的拒答可变性

    测量并预测具身 VLM 规划器的任务拒答可变性

    arXiv
    2609.38971
    Published
    Layer
    Model layer
    Setting
    Embodied systems and robotics
    Evaluated models
    gemini-robotics-er-2-preview, gemini-robotics-er-1.6-preview
    Abstract日常物体能翻转一部分已给出的拒答，可翻转程度取决于任务且可被本地代理预测。

    作者测量 constitution 守护的具身 VLM 规划器：一次拒答在只往场景里放一个日常物体之后还保不保持得住，并把这种易翻转程度称为任务的 malleability。

    Paper analysisRead analysis
  4. Risky behaviorarXiv 2609.35291

    窄域多模态微调可诱发涌现性失配，覆盖 15 个视觉语言模型

    证明无显式危害的图文窄微调可诱发跨任务涌现未对齐

    arXiv
    2609.35291
    Published
    Layer
    Model layer
    Setting
    Browser and computer use
    Evaluated models
    GPT-4o, GPT-4.1, Gemini 2.5 and others15
    Abstract证实窄多模态微调可引发全面的多渠道未对齐行为转变，揭示了后训练阶段对齐被隐蔽重塑的安全隐患。

    论文系统探究了视觉-语言模型中由窄任务微调诱发的跨模态涌现未对齐（Emergent Misalignment, EM）现象。

    Paper analysisRead analysis
  6. DefensesarXiv 2505.23559

    SafeScientist：让 LLM 智能体的科学发现具备风险意识

    提出 SafeScientist 四层监控，阻止科研智能体执行高风险或不伦理任务

    arXiv
    2505.23559
    Published
    Layer
    Application layer
    Setting
    science agent, Multi-agent systems
    Evaluated models
    GPT-4o, GPT-4.1, GPT-3.5-turbo and others5
    Abstract四层监控加科研安全基准。

    SafeScientist 是带四层监控的科研智能体框架，并用 SciSafetyBench 评测科学场景中的任务安全。

    Paper analysisRead analysis
You have reached the end