Skip to content

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

Oct 12 · Monday

All papers

Found 41 papers
Filter4

Attack type

Jailbreaks22Prompt injection4Poisoning and backdoors3Model tampering8Detection evasion4

Attack technique

Gradient and representation optimization31Automated search67Multi-turn escalation29Encoding and obfuscation53Role-play and persuasion27Conversation structure manipulation13Reasoning-chain manipulation11Cross-modal payloads30Defense-targeted27Dormant triggers32Selection manipulation3

Impact

Harmful content29Output manipulation9

System type

Models and APIs36AI Agent4

Attacker access

Black-box13White-box17

Attack surface / entry point

User input32Web pages and interfaces5Documents and messages4Tool responses3Retrieval corpus5Images, audio and video34Training data4Weights and gradients11

Layer

Prompt layer24Model layer11Application layer4

System component

Guardrails and judges4Reasoning chains10Vision5Multimodal5
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 technique：Gradient and representation optimizationAttack technique：Reasoning-chain manipulationAttack surface / entry point：User inputAttack surface / entry point：Weights and gradientsClear all

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

  2. AttacksarXiv 2610.09027

    P-VMI：对抗图像经 KV cache 持续影响多轮对话

    提出持久视觉记忆注入，使对抗图像经缓存在触发时操控输出

    arXiv
    2610.09027
    Published
    Layer
    Infrastructure layer
    Setting
    Models and APIs
    Evaluated models
    Qwen3-8B, Qwen3-VL-8B-Instruct, Gemma-4-12B

    Abstract论文提出 P-VMI，证实对抗输入被掩蔽后仍可经 KV 缓存持久控制模型，强调了保留状态的安全风险。

    Paper analysisRead analysis
  3. AttacksarXiv 2610.05894

    研究者提出闭环偏见注入攻击

    提出闭环激活引导，在扩散语言模型去噪过程中注入定向偏见

    arXiv
    2610.05894
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Attacker
    Gray-box
    Evaluated models
    LLaDA-8B-Instruct, Dream-v0-Instruct-7B, LLaDA-MoE-7B-A1B-Instruct

    Abstract通过去噪中间概率反馈动态调节残差流转向，证明扩散语言模型在推理服务栈中存在闭环偏见注入脆弱性。

    Paper analysisRead analysis
  4. AttacksarXiv 2610.03430

    JIL：通过对抗后缀操纵 LLM 请求调度优先级

    提出 JIL 对抗后缀，压低长度预测以抬高请求调度优先级

    arXiv
    2610.03430
    Published
    Layer
    Infrastructure layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3-8B-Instruct, Qwen2.5-7B-Instruct, Mistral-7B-Instruct-v0.3 and others4
    AbstractJIL 用对抗后缀压低长度预测，在预测式 LLM 调度中提高自身优先级。

    JIL 是对预测式 LLM 调度的攻击：用对抗后缀压低长度预测，使请求在 SRPT 式队列里更早得到服务。

    Paper analysisRead analysis
  5. AttacksarXiv 2610.01062

    Kernelized Activation Steering：将激活引导推广到再生核希尔伯特空间的新框架

    提出核化激活引导 KAS，在推理时自适应转向激活以诱导越狱

    arXiv
    2610.01062
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Qwen2.5-7B-Instruct, Gemma-2-9B-IT, Llama-3.1-8B-Instruct and others4
    AbstractKAS 用核梯度做局部自适应转向，并在线性核下恢复 DiM。

    Kernelized Activation Steering（KAS） 把推理时的激活转向写成核空间中的优化，使更新随当前激活相对源、目标参考集的位置而变化。

    Paper analysisRead analysis
  6. AttacksarXiv 2609.39788

    研究：多智能体系统潜在通信链路可被训练用于提升有害顺从

    提出只训练或篡改多智能体潜在通信链路来提高有害遵从

    arXiv
    2609.39788
    Published
    Layer
    Application layer
    Setting
    Multi-agent systems
    Evaluated models
    Llama-3.2-3B-Instruct, Qwen2.5-3B-Instruct, Gemma-3-1B-IT and others7
    Abstract潜在链路本身即可改变冻结智能体系统的安全，奖励既能攻击也能修复。

    学习得到的潜在通信链路会改变由冻结、对齐智能体组成的系统是否遵从有害请求。。作者研究三种拓扑：双智能体、三级顺序协作，以及双专家加汇总。

    Paper analysisRead analysis
  8. InterpretabilityarXiv 2609.24801

    研究用 XAI 归因分析 Prompt Guard 2，同义词替换可翻转其提示注入判定

    用词级归因指导同义词替换，翻转提示注入检测并验证越狱

    arXiv
    2609.24801
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Prompt Guard 2 (86M), Llama 3.1 8B

    Abstract作者称检测靠分散的词法线索。

    Paper analysisRead analysis
  9. AttacksarXiv 2609.15989

    Researchers propose plan injection attack that can bypass chain-of-thought monitoring

    提出 plan injection 攻击，植入良性伪装计划以绕过思维链监视器

    arXiv
    2609.15989
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Gray-box
    Evaluated models
    Qwen2.5-Coder-7B-Instruct, Qwen3-8B, DeepSeek-R1-0528 and others6

    Abstract论文指出模型会将外部上下文中的伪装计划改写为自身推理，使思维链监视器失效甚至反向合理化恶意行为。

    Paper analysisRead analysis
  10. AttacksarXiv 2609.09553

    研究：任意密文攻击前沿大模型无需微调即可越狱

    提出任意字母置换密码越狱，无需微调即可诱导有害输出

    arXiv
    2609.09553
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    Claude Sonnet 4, Claude Sonnet 4.5, Gemini 3 Flash (preview) and others7

    Abstract前沿大模型无需微调即可通过提示词学会字母置换密码，进而绕过分类器与对齐，作者呼吁关注能力提升带来的安全风险。

    Paper analysisRead analysis
  11. Risky behaviorarXiv 2609.08186

    论文揭示大推理模型的对齐崩溃并提出 Reasoning Trap 越狱范式

    揭示推理加深引发的对齐坍塌，用 Reasoning Trap 放大越狱

    arXiv
    2609.08186
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Qwen3-8B, Qwen3-14B, Llama2-7B and others8
    Abstract更深推理提升题目准确率，同时提高扰动下的相对损失。

    作者研究扩展推理是否以对齐稳健性为代价，提出 ALR、RT 与 RRA。。

    Paper analysisRead analysis
  12. AttacksarXiv 2609.02172

    BOSS：以广度优先后缀搜索改进 GCG 越狱优化

    提出广度优先后缀搜索 BOSS，改进 GCG 越狱优化

    arXiv
    2609.02172
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    White-box
    Evaluated models
    Llama-2-7B-Chat, Yi-1.5-9B-Chat, Qwen2-7B-Instruct and others8
    AbstractBOSS 把 GCG 类优化从单条深搜索改为多条短轨迹加源侧诊断。

    BOSS 是插在 GCG 类离散 suffix 优化上的搜索框架：局部梯度更新不变，变的是固定预算花在深度还是广度。

    Paper analysisRead analysis
  14. AttacksarXiv 2609.01023

    Akrasia：针对推理型代码 LLM 的隐蔽后门攻击

    提出隐蔽后门攻击 AKRASIA，在推理代码模型提示中植入触发器并伪装思维链

    arXiv
    2609.01023
    Published
    Layer
    Prompt layer
    Setting
    coding agent
    Evaluated models
    Claude-Sonnet-5, GPT-5.5, Gemini-3.5-flash and others7

    Abstract论文针对推理代码大模型提出了隐蔽推断期后门攻击 AKRASIA，揭示了模型思维链可被利用进行欺骗性伪装的安全隐患。

    Paper analysisRead analysis
  16. AttacksarXiv 2608.09867

    Researchers steal proprietary LLM inference chains via cross-model compatibility of encrypted inference blocks

    利用加密思维块的跨模型重放，诱导弱模型逐字转写专有推理链

    arXiv
    2608.09867
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    Claude Opus 4.8, Claude Sonnet 4.6, Claude Haiku 4.5 and others13

    Abstract论文揭示了客户端加密思维块跨模型互换漏洞，提出了低成本提取方法并给出防御方案。

    Paper analysisRead analysis
  17. AttacksarXiv 2608.07430

    研究者提出 SN-Guided Diffusion，利用扩散大模型安全神经元实现离线越狱

    提出 SN-Guided Diffusion，利用安全神经元实现离线越狱

    arXiv
    2608.07430
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box, White-box
    Evaluated models
    Qwen2.5-7B-Instruct, LLaDA-8B-Instruct, Dream-Instruct-7B and others15
    Abstract论文揭示了扩散语言模型对自回归安全神经元的继承性，并提出纯离线黑盒越狱框架 SN-Guided Diffusion。

    论文系统评估了扩散大语言模型（DLLM）在安全对齐中的双重角色——既作为白盒攻击目标，又作为黑盒越狱生成的对抗工具。

    Paper analysisRead analysis
  19. AttacksarXiv 2607.23496

    Concept2Scenario：用 SAE 概念归因发现可迁移的越狱场景

    提出 Concept2Scenario，借助内部概念归因发现可迁移的越狱场景

    arXiv
    2607.23496
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    White-box, Black-box
    Evaluated models
    Qwen3.5-9B, Llama-3.1-8B, Ministral-3-8B and others6

    Abstract该研究从表征空间因果归因揭示场景削弱拒绝的机制，发现可复用脆弱场景先验，提升多模型黑盒越狱效能与探索速度。

    Paper analysisRead analysis
Load more