Skip to content

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

Oct 12 · Monday

Agent safety · Papers

Found 2 papers
Filter2

Model risks

Dangerous agent actions2

Attack surface / entry point

User input4Web pages and interfaces3Documents and messages2Tool responses5
Select any value within a category; different categories must all match. Counts include the other filters you have selected.
Model risks：Dangerous agent actionsAttack surface / entry point：Documents and messagesTopic: Agent safetyClear all
  2. Evaluation and benchmarksarXiv 2601.21083

    OpenSec：测量对抗证据下事件响应 Agent 的校准

    提出 OpenSec，评测事件响应 Agent 在对抗证据下的遏制校准

    arXiv
    2601.21083
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    GPT-5.2, Claude Sonnet 4.5, Gemini 3 and others5
    AbstractOpenSec 以执行评分测 IR 智能体在对抗证据下是否过度遏制，作者称缺口在克制。

    OpenSec 用 dual-control 事件环境测量 IR 智能体在对抗证据下的校准：分数看实际遏制，不看报告文本。

    Paper analysisRead analysis
You have reached the end