Skip to content

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

Oct 12 · Monday

Defenses and guardrails · Papers

Found 5 papers
Filter2

Attack type

Jailbreaks5

Defense type

Detection and filtering5

Impact

Harmful content5

System type

Models and APIs4

Attacker access

Black-box5

Attack surface / entry point

User input5

Layer

Prompt layer4

System component

Guardrails and judges5Multimodal3
Select any value within a category; different categories must all match. Counts include the other filters you have selected.
Attacker access：Black-boxSystem component：Guardrails and judgesTopic: Defenses and guardrailsClear all
  1. DefensesarXiv 2610.03434

    Persona Guardrail：面向智能体系统的生产级防御框架

    提出 Persona Guardrail，强制客服智能体守住功能边界

    arXiv
    2610.03434
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Attacker
    Black-box
    Evaluated models
    Qwen3-30B-A3B, Qwen3-VL-30B, Qwen3-8B and others5
    AbstractPersona Guardrail 用功能边界做同步输入输出校验，并在 PAGE 与生产延迟预算下给出评测。

    Persona Guardrail 是客服向智能体的同步运行时防御：先写明功能边界，再决定放行或拦截，而不是只检测某一攻击类。

    Paper analysisRead analysis
  2. DefensesarXiv 2610.02853

    研究提出基于收缩约束状态空间模型的有界可达性与越狱检测方法

    提出收缩约束 SSM 安全头，认证嵌入扰动并检测越狱

    arXiv
    2610.02853
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    White-box, Black-box
    Evaluated models
    ToySSMClassifier, S4 safety head
    Abstract收缩给出玩具 LTI 的有界认证。

    给 SSM 安全头一个收缩条件，并在越狱检测上使用收缩正则的 S4 头。。

    Paper analysisRead analysis
You have reached the end