Skip to content

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

Oct 13 · Tuesday

All papers

Found 2 papers
Filter5

Attack type

Jailbreaks8Prompt injection33Poisoning and backdoors14Extraction and theft2Detection evasion10Malicious use3

Defense type

Model hardening3Inference-time intervention5Detection and filtering2Watermarking and provenance3

Layer

Model layer0Application layer2Infrastructure layer0
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 215papers have no tags yet and are not included in filtered results.
Attack type：Extraction and theftDefense type：Detection and filteringLayer：Model layerLayer：Application layerLayer：Infrastructure layerClear all

Another 215papers do not have category tags yet and are not included in filtered results.

  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
You have reached the end