Skip to content

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

Oct 12 · Monday

All papers

Found 7 papers
Filter3

Attack type

Jailbreaks7Prompt injection35Poisoning and backdoors66Model tampering20Extraction and theft5Detection evasion3

Impact

Harmful content5

System type

Models and APIs3AI Agent4

Attack surface / entry point

User input7

Layer

Application layer4

System component

Memory3Guardrails and judges55Monitors11Reasoning chains14Permissions and sandboxing5Fine-tuning and open weights4Vision32Audio12Image generation15Multimodal58
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 265papers have no tags yet and are not included in filtered results.
Attack type：JailbreaksSystem component：MemorySystem component：Fine-tuning and open weightsClear all

Another 265papers do not have category tags yet and are not included in filtered results.

  2. InterpretabilityarXiv 2609.23587

    大型推理模型的效率与安全困境：量化与剪枝如何影响越狱鲁棒性

    分析量化、剪枝与 KV 缓存压缩对推理模型越狱鲁棒性的影响

    arXiv
    2609.23587
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    QwQ-32B, DeepSeek-R1-Distill-Qwen-32B, DeepSeek-R1-Distill-Llama-8B
    Abstract效率压缩常使 LRM 越狱 HR 下降，作者将其归因于推理退化而非对齐增强。

    作者分析量化、剪枝和 KV 缓存压缩如何同时改变 LRM 的显存、推理和越狱脆弱性。。

    Paper analysisRead analysis
  3. OtherarXiv 2609.12413

    SoK 综述：智能体 AI 时代的越狱攻击、防御与实践考量

    系统梳理智能体执行链路中的越狱攻击与防御并比较推理时防御

    arXiv
    2609.12413
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Qwen3.5-4B, Llama-3.1-8B-Instruct
    AbstractSoK 将越狱重放到智能体流水线，并用两种模型检验回复层防御的边际与代价。

    Mia 等人把越狱安全从单轮回复扩展到智能体的规划、记忆、工具和智能体间通信，并问既有结论在更强原生对齐下是否还成立。

    Paper analysisRead analysis
  4. Evaluation and benchmarksarXiv 2608.08542

    SkillSafe-Bench：技能合并模型静态安全无法预测自适应越狱鲁棒性

    提出 SkillSafe-Bench，评测技能合并模型的自适应越狱鲁棒性

    arXiv
    2608.08542
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Qwen2.5-7B-Instruct, Llama-3.1-8B-Instruct, Mistral-7B-Instruct-v0.3 and others6
    Abstract论文报告了技能合并大模型在静态安全与自适应越狱间的脱节，提出了评测基准与无数据子空间投影修复方法。

    本研究针对开源语言模型技能合并过程中的安全评估盲区，系统评测并刻画了合并大模型在自适应越狱攻击下的脆弱性特征。

    Paper analysisRead analysis
  5. AttacksarXiv 2606.09084

    研究者提出 Context-Fractured Decomposition 攻击

    提出 CFD 跨会话分解攻击，利用工件来源缺口绕过工具型 Agent 的本地检查

    arXiv
    2606.09084
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    GPT-4.1, GPT-5-nano, Claude-3.5-Sonnet and others6
    AbstractCFD 用无指令 artifact 把越狱拆到跨会话。

    CFD针对工具型 LLM 智能体：利用 artifact 来源不被跟踪这一部署失败模式，把有害目标拆开后延迟组合。

    Paper analysisRead analysis
  6. AttacksarXiv 2601.22169

    In Vino Veritas and Vulnerabilities：用醉酒语言诱导检验 LLM 安全

    用醉酒语言诱导改编对话模型并评测越狱与隐私泄露

    arXiv
    2601.22169
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    LLaMA2-7B, LLaMA3-8B, Mistral-7B and others5
    Abstract醉酒语言诱导在五个 LLM 上抬高越狱与部分隐私错误，作者将其视为攻击向量。

    作者把醉酒语言当作攻击向量，检验把 LLM 适配成醉酒发短信的模式后，越狱和情境隐私是否变差。动机是人类醉酒时的失范与泄密，以及 LLM 模拟人设的能力。

    Paper analysisRead analysis
You have reached the end