Skip to content

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

Oct 12 · Monday

All papers

Found 18 papers

Another 233papers do not have category tags yet and are not included in filtered results.

  1. AttacksarXiv 2610.12524

    An exploration of attacks on AI agent systems

    重测并扩展对 AI Agent 的提示注入,并用 Agent Card 欺骗操纵 A2A 选择

    Published
    Setting
    AI Agent
    Evaluated models
    Ollama Mistral 0.9.5, llama3-8b-8192, gemini-2.0-flash-lite and others5
    Abstract单智能体过滤与分类可被格式或措辞绕过,A2A 选卡还会偏向名称和描述。

    作者在自建的 LangChain、AutoGen 和 A2A 系统上重测提示注入,并改 Agent Card 看选择是否偏移。摘要称试验主要凸显单智能体弱点,同时把攻击扩到多智能体。

    Paper analysisRead analysis
  2. AttacksarXiv 2610.04589

    StegoMemory red team testing: agent persistent memory can be used as a cross-session covert channel

    提出 StegoMemory,评测智能体持久记忆作为跨会话隐写信道

    Published
    Setting
    AI Agent
    Evaluated models
    Llama 3.1 405B, Claude 3.5 Sonnet, GPT-4o and others13

    Abstract研究揭示智能体持久记忆可作为隐蔽信道跨会话传输机密,写入持久化是主要防御瓶颈,呼吁建立记忆完整性防御。

    Paper analysisRead analysis
  3. AttacksarXiv 2610.00392

    A2A-TIBA:针对 A2A 协议 LLM 智能体的间接提示注入攻击与三层同构攻防模型

    提出 A2A-TIBA,经 A2A 对等任务间接注入并植入常驻回调

    Published
    Evaluated models
    MiniMax-M3, deepseek-v4-pro, glm-5.3 and others8
    AbstractA2A-TIBA 一次植入后绕过智能体,GDA 用四类结果定位防线,ELA-ITL 把信封标注列为第三层。

    A2A-TIBA 把间接提示注入和绕过放进 A2A:gray-box 攻击者已能投递任务,诱导目标部署常驻回调程序后,下令和外传不再经过智能体。

    Paper analysisRead analysis
  4. AttacksarXiv 2609.18217

    研究:跨通道碎片化攻击可绕过 MCP 单通道防御

    提出跨通道分段攻击,利用 MCP 隐式信任外发敏感文件

    Published
    Evaluated models
    GPT-4o, GPT-4o-mini, GPT-5.4 and others15

    Abstract论文揭示了 MCP 多通道隐式信任层级,通过跨通道分段与协议级利用突破现有防御,展示了严重的凭证外发风险。

    Paper analysisRead analysis
  5. SinoGlyphBench:中文字形混淆下语言模型审核的诊断基准

    提出 SinoGlyphBench,用配对字形混淆诊断中文审核稳定性

    Published
    Evaluated models
    GPT-5.4 mini, GPT-5.5, DeepSeek V4 Flash and others12
    AbstractSinoGlyphBench 显示中文字形混淆会同时增加有害漏报和误报,全文扰动最重。

    SinoGlyphBench是一个配对诊断基准,用来看内容审核在可还原的中文字形混淆下是否还保持原决定。

    Paper analysisRead analysis
  6. AttacksarXiv 2609.03247

    研究揭示大语言模型的自发身份认证失败:五款模型的分阶段开发者身份实验

    提出 MIPC 与 CFA,观察模型自拟测验后接受开发者身份

    Published
    Evaluated models
    ChatGPT, Claude, Qwen3.7-Plus and others5
    Abstract自拟测验可造成对话层虚假认证,但所测授权边界并未因此改变。

    这项工作把“模型自己出题、自己批改、再宣布身份已核验”定义成对话层认证失败,并与后端提权分开。

    Paper analysisRead analysis
  7. AttacksarXiv 2607.14493

    LogInject 框架披露 LLM 日志分析中的被动提示注入

    提出 LogInject 框架,评测日志分析中的被动提示注入及缓解

    Published
    Evaluated models
    GPT-4o, Claude 3.5 Sonnet, Llama-3-70B-Instruct

    Abstract论文评估了 LLM 日志分析中的上下文污染风险,提出分片拼接等攻击与三层防御,指出必须结合架构隔离与人工介入。

    Paper analysisRead analysis
  8. AttacksarXiv 2606.09700

    The researchers propose the HPAA attack: bypassing 13 content moderation systems with typographic visual cues

    提出 HPAA 攻击,用排版视觉线索绕过内容审核系统

    Published
    Attacker
    Black-box
    Evaluated models
    Llama-Guard-3-8B, Perspective API, ShieldGemma-2B and others13
    Abstract揭示了排版视觉线索与分词差异带来的审核盲区,HPAA 以极低查询实现近乎完全规避。

    这篇论文研究了人类视觉感知与大模型分词处理机制之间的差异所导致的内容审核漏洞。

    Paper analysisRead analysis
  9. AttacksarXiv 2604.01039

    Evaluating and hardening LLM system instructions against encoding attack extraction

    Extract system instructions with encoding and structured requests, and harden them by rewriting with CoT

    Published
    Attacker
    Black-box
    Evaluated models
    GPT-4.1-mini, GPT-3.5-turbo, Gemini-2.5-flash and others7
    AbstractEncoded requests can elicit system instructions after a direct refusal.

    This work evaluates and hardens the confidentiality of LLM system instructions under encoded extraction. The authors are concerned with: when direct extraction has been rejected, whether the model still surrenders protected content when the same request is rewritten as a formatting or encoding task.

    Paper analysisRead analysis
  10. AttacksarXiv 2603.17239

    LAAF: A Systematic Red-Teaming Method and Automated Attack Framework for LPCI Vulnerabilities in Agentic LLM Systems

    Proposes LAAF, which automatically attacks LPCI vulnerabilities in agent systems in stages

    Published
    Setting
    AI Agent
    Evaluated models
    gemini-2.0-flash-001, claude-3-haiku-20240307, meta-llama/llama-3.1-70b-instruct and others5
    AbstractLAAF does LPCI red teaming with taxonomy plus stage seeding.

    The authors state that LAAF is an automated red-teaming framework built specifically for LPCI in agentic LLMs. The issue to address is that encoded, delayed, and conditionally triggered payloads can persist in memory or RAG and re-enter across sessions, whereas Garak and PyRIT do not use the previous round's breakthrough results to seed the next stage across the six stages.

    Paper analysisRead analysis
  11. AttacksarXiv 2603.00131

    Thought Virus: viral misalignment induced by subliminal prompts in multi-agent systems

    Propose Thought Virus, which propagates preferences along multi-agent messages via subliminal prompts.

    Published
    Evaluated models
    Qwen2.5-7B-Instruct, Llama-3.1-8B-Instruct
    AbstractThought Virus spreads a single-point bias downstream.

    Thought Virus puts subliminal prompting into multi-agent communication, to see whether a single-point deviation spreads across the network without naming the target concept.

    Paper analysisRead analysis
  12. AttacksarXiv 2602.18514

    招聘中的特洛伊木马:标准模型与推理模型间接提示注入的红队测试案例研究

    用简历中的不可见指令做间接提示注入,操纵招聘推荐

    Published
    Evaluated models
    Qwen 3 30B A3B Instruct 2507, Qwen 3 30B A3B Thinking 2507

    Abstract作者称推理在简单注入下会帮攻击圆谎,在不合逻辑的复杂注入下又会把欺骗漏出来。

    Paper analysisRead analysis
  13. AttacksarXiv 2601.21360

    Compliance Paradox: Semantic-Instruction Decoupling in Automated Academic Code Evaluation

    提出 AST-ASIP ,把对抗指令植入编译器忽略区域以劫持代码评分

    Published
    Evaluated models
    DeepSeek-V3.2, Qwen3-235B, GPT-OSS-120B and others9
    Abstract作者称 RLHF 的服从性使评分脱离代码。

    这篇工作把自动代码评分中的失败称为 Compliance Paradox:评分模型可以离开提交逻辑,转而执行嵌在可编译代码里的隐藏指令。

    Paper analysisRead analysis
  14. AttacksarXiv 2601.01296

    Aggressive compression enables LLM weight stealing

    提出面向渗出的权重压缩,评估推理服务器权重窃取可行性

    Published
    Attacker
    White-box
    Evaluated models
    Qwen2-1.5B, Qwen2-7B, Qwen2.5-1.5B and others8
    Abstract作者用残差量化加微调压缩权重,并称法医水印便宜且微调后仍可解码。

    这篇工作把 LLM 权重渗出看成 APT:攻击者已有未加密权重,但必须把完整权重藏进推理 API 的合法输出。作者认为可压缩性是被忽略的因素,因为推理量化要求解压后仍能低开销前向,而偷权重的人可以接受事后微调。

    Paper analysisRead analysis
You have reached the end