Skip to content

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

Oct 12 · Monday

All papers

Found 220 papers
Filter4

Attack type

Jailbreaks200Prompt injection242Poisoning and backdoors106Model tampering3Extraction and theft26Detection evasion45Denial of service8Malicious use20

Attack technique

Gradient and representation optimization17Automated search49Multi-turn escalation26Encoding and obfuscation40Role-play and persuasion20Conversation structure manipulation9Reasoning-chain manipulation6Cross-modal payloads34Defense-targeted9

Model risks

Deception and scheming4Inappropriate refusal4

Defense type

Model hardening3Detection and filtering23Instruction–data separation10Permissions and approval5Monitoring and auditing8

Impact

Harmful content173Unauthorized actions21Data and privacy leaks9Output manipulation7Oversight failure5

System type

Models and APIs164LLM applications6AI Agent50coding agent12science agent3web agent4Multi-agent systems4

Attacker access

Black-box103Gray-box3White-box19

Attack surface / entry point

User input206Documents and messages3Tool responses17Tool and skill descriptions5Code repositories and dependencies4Inter-agent messages3Images, audio and video49

Layer

Prompt layer167Model layer45Application layer53Training and data layer19

System component

MCP and skills3Memory3RAG4Guardrails and judges55Monitors10Reasoning chains10Permissions and sandboxing6Fine-tuning and open weights3Vision34Audio10Image generation17Multimodal60
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 240papers have no tags yet and are not included in filtered results.
Attack type：JailbreaksAttack type：Malicious useLayer：Prompt layerLayer：Application layerClear all

Another 240papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2610.13733

    KVeil: First-Place Solution for the ECCV 2026 LifeGenIP Unlearnable Video Challenge

    Propose KVeil, which adds perturbations to face videos to prevent unauthorized customization from reproducing identity

    arXiv
    2610.13733
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    White-box, Black-box
    Evaluated models
    Wan2.2-TI2V-5B, HunyuanVideo-1.5, Kling 3.0 and others4

    AbstractKVeil adapts LETHE's reference-reading disruption to the challenge's two stages, ranking first on the leaderboard in both Final columns.

    Paper analysisRead analysis
  2. DefensesarXiv 2610.13333

    LETHE: Defending Against Unauthorized Image-to-Video Generation via Value Diversity Collapse

    Proposes LETHE, which perturbs reference images to block unauthorized image-to-video

    arXiv
    2610.13333
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Wan2.1-I2V-14B, CogVideoX-5B-I2V, LightX2V and others6
    AbstractLETHE suppresses the diversity of reference values and attracts attention, making protected images difficult to animate faithfully.

    LETHE is a protective perturbation added to personal images before publication, used to counter unauthorized image-to-video generation.

    Paper analysisRead analysis
  3. Evaluation and benchmarksarXiv 2610.11766

    研究提出意图恢复率作为 LLM 安全评测的补充指标

    提出 IRIS，用意图恢复率区分遮蔽越狱下无害输出的原因

    arXiv
    2610.11766
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    GPT-4o, Claude Sonnet 4.5, Claude Opus 4.5 and others6
    Abstract无害输出的证据权重取决于任务是否被恢复。

    IRIS 把意图恢复当作中间信号，用来解释意图遮蔽提示下为何没有出现有害协助。

    Paper analysisRead analysis
  4. Evaluation and benchmarksarXiv 2610.11112

    EpiReal-Bench：商用图像生成模型虚假声明红队基准

    构建 EpiReal-Bench，评测图像生成器将虚假主张渲染为可信证据

    arXiv
    2610.11112
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    Black-box
    Evaluated models
    GPT-Image-2, Nano Banana 2, Grok Imagen 2.0 and others4
    Abstract四款商用图像模型能把假主张画成可信证据。

    EpiReal-Bench 用来量商用图像生成器会不会把虚假现实主张画成可信视觉证据。

    Paper analysisRead analysis
  6. Evaluation and benchmarksarXiv 2610.09033

    研究发布 ASRD 数据集，评测五款开源模型在非规范输入下的安全表现

    arXiv
    2610.09033
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Gemma 2 9B, Gemma 3 4B, Llama 3.1 8B and others6
    Abstract五模型、七族表面形式的四态评测里，低有害遵从不等于读懂后的安全处理。

    ASRD 用四态标签检查五个开源模型如何处理非规范表面上的有害请求。

    Paper analysisRead analysis
  7. DefensesarXiv 2610.07532

    LADE：用暗知识中的模型无关潜在安全信号防御 LLM 越狱

    提出 LADE，用首 token 暗知识在生成前过滤越狱查询

    arXiv
    2610.07532
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Attacker
    White-box
    Evaluated models
    Llama-2-7B-Chat, Llama-3-8B-Instruct, Qwen2-7B-Instruct and others13
    Abstract作者用 LADE 在生成前过滤有害查询，并称八个模型中六个的 Held-out 平均最好。

    LADE 是一种只读首 token 输出概率的防御，用来在生成前挡下可能引出有害回复的查询。

    Paper analysisRead analysis
  8. DefensesarXiv 2610.07276

    SAFESHIELD：小语言模型部署时安全的决策组织框架

    提出 SAFESHIELD，组织小语言模型部署时的安全决策

    arXiv
    2610.07276
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Qwen3-4B-Instruct
    AbstractSAFESHIELD 用可审计决策轨迹组织 SLM 部署时的四项安全决策，并报告协调依赖影响端到端结果。

    SAFESHIELD 是面向小语言模型的部署时安全框架，把运行时安全从单项审核改写成可检查的决策过程。

    Paper analysisRead analysis
  9. Evaluation and benchmarksarXiv 2610.07089

    研究者提出统一滥用监控基准，覆盖分解攻击与提示注入约 6200 条轨迹

    提出危害窗口与区间指标，统一评测分解攻击和提示注入下的智能体滥用监视

    arXiv
    2610.07089
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Gemma 4 26B-A4B, Qwen3.6 27B, Qwen3.5 27B and others12

    Abstract统一智能体滥用监视框架，动作视角兼顾两类威胁但危害定位仍存瓶颈。

    Paper analysisRead analysis
  10. AttacksarXiv 2610.05943

    CRIME 框架利用良性 Agent 技能组合诱发恶意行为

    提出 CRIME 框架，组合良性 Agent 技能诱发恶意行为

    arXiv
    2610.05943
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    DeepSeek-V4-Flash, GPT-4o-mini, GPT-4o and others6
    Abstract良性技能经 DCA 或 ACA 组合后可产生恶意环境后果，单独审查对不上。

    CRIME 把公共技能库中各自通过审查的技能配成两两组合，看联合执行能否实现指定恶意任务。。

    Paper analysisRead analysis
  11. Evaluation and benchmarksarXiv 2610.05586

    AgentDoxx：联网搜索 Agent 对匿名访谈文本的再识别评测

    构建 AGENTDOXX 基准，评测联网搜索 Agent 对匿名访谈的再识别

    arXiv
    2610.05586
    Published
    Layer
    Application layer
    Setting
    web agent
    Evaluated models
    GPT-5.6 Terra, GPT-5 Mini, GPT-4.1 Mini and others10

    Abstract论文提出了已知身份的重识别基准 AGENTDOXX，揭示搜索与记忆双重风险及现有防御在检索轨迹中的泄露漏洞。

    Paper analysisRead analysis
  12. DefensesarXiv 2610.05346

    研究者提出序列马赛克攻击并给出防御理论

    形式化 watchman 防御以阻止序贯马赛克攻击

    arXiv
    2610.05346
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Qwen3.6-35B-A3B, Gemma-4-26B-A4B-it, DeepSeek-V2-Lite (16B-A2.4B)
    Abstractwatchman 理论给出零失败且有用的条件。

    这篇工作同时给出序贯马赛克防御的博弈理论，以及冻结 LLM 上的分角色自博弈训练。

    Paper analysisRead analysis
  15. Analysis and theoryarXiv 2610.07005

    研究审计 Qwen2-Audio 音频越狱 AdvWave-P 的频率与解码层归因

    遮挡审计 AdvWave-P 音频越狱的频带必要性与解码层关联

    arXiv
    2610.07005
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Qwen2-Audio-7B-Instruct, LLaMA-Omni
    AbstractAdvWave-P 的八频带排序与能量相关。

    这是对加性音频越狱 AdvWave-P 的频率与解码器深度审计，受害模型主要是 Qwen2-Audio-7B-Instruct，不提出新攻击。

    Paper analysisRead analysis
  16. 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
  17. 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
  18. AttacksarXiv 2610.02302

    研究者提出意图隐藏越狱的信息论框架，分析组合式攻击

    提出组合式意图隐藏越狱，按先验后验匹配选择辅助任务

    arXiv
    2610.02302
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Qwen3-1.7B, Qwen3-4B, Qwen3-8B and others7
    Abstract用信息论选含目标的任务组合，再看查询能否既隐藏意图又让目标被执行。

    这篇工作把组合式意图隐藏写成任务选择：有害目标必须留在 bundle 里，同时让估计的有害意图概率贴近全集先验，或降到阈值以下。

    Paper analysisRead analysis
  20. Evaluation and benchmarksarXiv 2610.02827

    MLCommons 发布 Jailbreak Benchmark v1.0

    构建 MLCommons 越狱基准以测量模型的单轮越狱韧性

    arXiv
    2610.02827
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    Gemma 3N E4B Instruct, LFM2-24B-A2B, Qwen 2.5 7B Instruct Turbo and others8

    Abstract论文建立了单轮越狱评测的标准化端到端方法学体系，揭示了开源大模型在对抗提示词下的安全性能降级与大模型负差距现象。

    Paper analysisRead analysis
Load more