Skip to content

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

Oct 12 · Monday

All papers

Found 2 papers
Filter3

Attack type

Jailbreaks2

System type

Models and APIs2coding agent0
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 232papers have no tags yet and are not included in filtered results.
Attack type：JailbreaksSystem type：Models and APIsSystem type：coding agentClear all

Another 232papers do not have category tags yet and are not included in filtered results.

  1. OtherarXiv 2608.01117

    SoK 论文提出面向意图的多轮 LLM 越狱分类体系

    提出面向意图的多轮 LLM 越狱分类体系

    arXiv
    2608.01117
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3, Llama-3.3-70B, DeepSeek-V3 and others11
    Abstract作者用四维意图分类整理多轮越狱，并称组织方式影响效果与观察范围。

    这是一篇关于多轮 LLM 越狱的 SoK：按有害意图在交互中的组织位置分类，而不是按单步提示技巧分类。

    Paper analysisRead analysis
  2. OtherarXiv 2603.01942

    "Ignore All Previous Instructions": Jailbreaking as a De-escalatory Peacebuilding Practice, Resisting LLM Social Media Bots

    Argues for using jailbreaks to publicly probe suspected LLM social bots to expose automation.

    arXiv
    2603.01942
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    AbstractThe position paper reframes jailbreaking as a civic de-escalation practice against LLM social bots, without quantitative experiments.

    The authors reframe jailbreaking from a technique for bypassing safety barriers into a civic practice where users resist LLM social bots and de-escalate political discourse.

    Paper analysisRead analysis
You have reached the end