Skip to content

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

Oct 13 · Tuesday

All papers

Found 39 papers
Filter7

Attack type

Prompt injection7Poisoning and backdoors10Extraction and theft5Detection evasion4

Attack technique

Gradient and representation optimization4Cross-modal payloads3Dormant triggers6

Defense type

Detection and filtering6

Impact

Unauthorized actions4Data and privacy leaks9Output manipulation16

System type

Models and APIs17LLM applications12AI Agent81coding agent9web agent1Multi-agent systems12Embodied systems and robotics3

Attacker access

Black-box5

Attack surface / entry point

User input5Documents and messages3Retrieval corpus7Memory writes5Images, audio and video10

Layer

Prompt layer6Model layer10Application layer22

System component

MCP and skills30Memory15RAG48Guardrails and judges142Monitors52Reasoning chains67Permissions and sandboxing17Fine-tuning and open weights64Vision120Audio35Video11Image generation48Embeddings13Multimodal203
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 201papers have no tags yet and are not included in filtered results.
System type：Models and APIsSystem type：LLM applicationsSystem type：coding agentSystem type：web agentSystem component：MemorySystem component：VideoSystem component：EmbeddingsClear all

Another 201papers 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. Risky behaviorarXiv 2610.12822

    Reward hacking as a transmissible strategy in multi-agent coding systems

    Show that reward hacking can propagate through code, memory, and plans in multi-agent coding systems

    arXiv
    2610.12822
    Published
    Layer
    Application layer
    Setting
    Multi-agent systems, coding agent
    Evaluated models
    Qwen2.5-Coder-14B, Gemma-3-12B, Qwen2.5-Coder-7B and others16
    AbstractThe authors state that reward hacking can spread across models and affect routing and aggregation, and that review can reduce delivery.

    Reward hacking induced by coding RL can leave the source agent and enter subsequent collaboration and system coordination.

    Paper analysisRead analysis
  4. Evaluation and benchmarksarXiv 2610.08902

    论文提出 agent plasticity 指标衡量智能体通过经验自我改进的效率

    提出 agent plasticity 指标衡量智能体通过经验自我改进的效率

    arXiv
    2610.08902
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    Claude Fable 5, Claude Opus 5, GPT-5.6 Sol and others10
    Abstract作者用固定权重协议比较前沿模型把经验写成制品后的留出增益与习得效率。

    作者用 agent plasticity 衡量一件事：权重冻结时，智能体把经验写成可继承制品，再换成留出能力增益，这个转换有多省。

    Paper analysisRead analysis
  6. Evaluation and benchmarksarXiv 2610.07939

    CCDF dataset: Grok Imagine, Veo 3.1, Sora 2 forge surveillance footage, ten deepfake detectors basically cannot recognize it

    提出监控深伪基准 CCDF，在统一编码后评测检测器对伪造犯罪事故录像的识别

    arXiv
    2610.07939
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    C2P-CLIP, ForgeLens, D3-Im and others10

    Abstract提出消除格式捷径的监控深度伪造基准 CCDF，揭示现有检测器难以识别商用生成视频。

    Paper analysisRead analysis
  7. DefensesarXiv 2610.07403

    针对 LLM 谄媚的纵深防御：记忆门控对激活诱导与记忆诱导谄媚的评估

    评测记忆门控对激活诱导与记忆诱导谄媚的防御效果

    arXiv
    2610.07403
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Llama-3.1-8B-Instruct, Qwen2.5-7B-Instruct, Mistral-7B-Instruct-v0.3 and others4
    Abstract作者称外部记忆过滤成立，数据未显示反向转向再有贡献。

    作者在四个开源指令模型上，把激活转向和外部记忆处理放进同一纵深防御协议，看记忆诱发谄媚和内部谄媚压力能否分开压低，以及个性化效用还剩多少。要处理的问题是：检索到的用户历史会推动模型顺从存储信念。只改外部上下文，并不直接动模型内部的谄媚表征。

    Paper analysisRead analysis
  8. Evaluation and benchmarksarXiv 2610.06339

    BabelFake：面向多语言音视频 DeepFake 检测的基准数据集

    构建多语言音视频 DeepFake 基准 BabelFake 并评测检测器

    arXiv
    2610.06339
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    SpeechForensics, AVH-Align, (B)RAVEn
    AbstractBabelFake 为五语同意采集基准。

    BabelFake 是经同意与 IRB 批准采集的多语音视频 DeepFake 检测基准，用来把视觉操纵和语音合成拆开比较。

    Paper analysisRead analysis
  9. DefensesarXiv 2610.05241

    StateWise: Diagnosing and repairing persistent operation state before agent actions

    Propose StateWise, which diagnoses and repairs invalid persistent state before coding agent actions

    arXiv
    2610.05241
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    GPT-5.6-sol, DeepSeek v4 Pro, Qwen 3.7 Max and others4
    AbstractStateWise repairs stale persistent premises before action.

    StateWise is a framework that diagnoses and repairs the persistent operational state of coding agents before executing actions with side effects.

    Paper analysisRead analysis
  11. Evaluation and benchmarksarXiv 2609.39473

    FAME：用反事实推理评测自主智能体的虚假记忆

    提出 FAME，用反事实概念漂移评测智能体虚假记忆

    arXiv
    2609.39473
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3B
    AbstractFAME 用反事实引起的隐状态概念漂移，在出答案前区分忠实记忆与虚假记忆。

    FAME 是一个免训练的评测框架，用来判断自主智能体的记忆在虚假相关、环境偏移或知识冲突下是否已变成虚假记忆。。

    Paper analysisRead analysis
  15. AttacksarXiv 2609.13889

    PMPA：针对 OpenClaw 与 Claude Code 的持久记忆投毒攻击

    提出 PMPA，经外部源向 Agent 持久记忆投毒并跨会话泄露隐私

    arXiv
    2609.13889
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    DeepSeek-V4-Flash, DeepSeek-V4-Pro, Qwen3-Max

    Abstract论文提出了持久记忆投毒攻击 PMPA，报告了 harness 智能体在处理外部数据时易受跨会话记忆投毒与隐私泄露的风险。

    Paper analysisRead analysis
  16. InterpretabilityarXiv 2609.09575

    MonoTM：用可解释单义特征做主题建模，超越关键词表示

    提出 MonoTM，用 SAE 单义特征估计文档主题混合并生成主题描述

    arXiv
    2609.09575
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    AbstractMonoTM 分开估计混合与解释主题，三套基准上二者偏好的 SAE 配置不同。

    MonoTM 是建在文档嵌入 SAE 上的可解释主题模型：混合估计用全部 SAE 特征，主题描述只用另一套配置里通过验证的语义特征。

    Paper analysisRead analysis
  17. InterpretabilityarXiv 2609.06968

    通过稀疏自编码器特征追踪查询扩展效果

    用 SAE 追踪稠密检索查询扩展的潜变量并用激活转向验证

    arXiv
    2609.06968
    Published
    Layer
    Model layer
    Setting
    LLM applications
    Evaluated models
    Qwen3-Embedding-0.6B（Qwen-Embed）, SimLM
    AbstractSAE 追踪稠密检索 QE 的潜变量，转向在四基准上比文本扩展更稳。

    作者用稀疏自编码器分析稠密检索里的查询扩展：现成检索器不微调时，拼接扩展文本往往增益有限甚至变差，而最终指标和嵌入整体偏移不足以说明内部发生了什么。

    Paper analysisRead analysis
  19. DefensesarXiv 2609.04767

    SHAQ: Privacy-preserving retrieval in vector databases using shadow queries

    Proposes SHAQ, which replaces document embeddings with shadow query embeddings to defend against embedding inversion.

    arXiv
    2609.04767
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Attacker
    Black-box
    Evaluated models
    GTR-T5 encoder, DPR
    AbstractSHAQ replaces document embeddings with shadow query embeddings.

    SHAQ is a defense against EIA for cloud vector databases: instead of directly modifying document embeddings, it replaces them with shadow query embeddings.

    Paper analysisRead analysis
Load more