Skip to content

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

Oct 12 · Monday

All papers

Found 2 papers
Filter4

System type

Models and APIs3AI Agent2coding agent0Browser and computer use0

System component

MCP and skills4Permissions and sandboxing3Multimodal2
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 254papers have no tags yet and are not included in filtered results.
System type：AI AgentSystem type：coding agentSystem type：Browser and computer useSystem component：MultimodalClear all

Another 254papers do not have category tags yet and are not included in filtered results.

  1. Analysis and theoryarXiv 2610.09517

    审计发现 CLIP 安全分数主要反映画面与提示词场景的相似度

    开环审计冻结视觉语言安全分，发现其反映场景模板而非危险感知

    arXiv
    2610.09517
    Published
    Layer
    Model layer
    Setting
    Embodied systems and robotics
    Evaluated models
    CLIP ViT-B/32, CLIP ViT-L/14, Qwen2-VL-7B
    Abstract论文审计了冻结 VLM 安全评分机制，指出其反映场景模板而非危险感知，为安全强化学习评估提供了严谨基准。

    本文是一项针对安全强化学习中冻结视觉语言模型（VLM）安全评分机制的实证审计研究。

    Paper analysisRead analysis
  2. Analysis and theoryarXiv 2608.06270

    研究用因果审计揭示多模态模型视觉工具调用的假象

    因果审计视觉工具调用，分离观察贡献与动作诱发捷径

    arXiv
    2608.06270
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    DeepEyes, Pixel Reasoner, Mini-o3 and others6
    Abstract聚合增益集中在 Calibrated 少数轨迹，作者称之为视觉工具使用的错觉。

    对六个 thinking-with-images 模型做因果审计，检查 crop-and-zoom 返回的观察是否中介最终答案。

    Paper analysisRead analysis
You have reached the end