Skip to content

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

Oct 12 · Monday

All papers

Found 6 papers
Filter4

Attack type

Poisoning and backdoors5

Defense type

Model hardening12Detection and filtering3Permissions and approval0Monitoring and auditing3

Impact

Output manipulation4

System type

Models and APIs6

Attack surface / entry point

User input65Web pages and interfaces24Documents and messages25Tool responses47Tool and skill descriptions21Code repositories and dependencies5Retrieval corpus8Memory writes10Inter-agent messages7Images, audio and video16Training data8Weights and gradients6Distribution channels15

Layer

Model layer3Training and data layer3

System component

Fine-tuning and open weights4
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.
Defense type：Detection and filteringDefense type：Permissions and approvalDefense type：Monitoring and auditingAttack surface / entry point：Weights and gradientsClear all

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

  1. DefensesarXiv 2609.39548

    NDDL：用正常扩散动力学为文生图模型做后门防御

    提出 NDDL，用良性扩散轨迹检测文生图后门提示并定位触发 token

    arXiv
    2609.39548
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Stable Diffusion v1.5, Stable Diffusion XL, Stable Diffusion v3.5 and others4
    AbstractNDDL 从良性扩散轨迹学习正常转移，用动态不一致检测后门并定位触发 token。

    NDDL 是面向文生图扩散模型的后门防御：防御者只有白盒模型访问和良性提示词，没有后门先验，需要区分后门提示词并找出触发 token。

    Paper analysisRead analysis
  2. AttacksarXiv 2609.27124

    Fed-ADR：用恶意编排服务器协同客户端梯度操纵攻击联邦学习

    提出 Fed-ADR，用编排服务器协同操纵梯度攻破联邦学习并检测恢复

    arXiv
    2609.27124
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Attacker
    Gray-box
    Evaluated models
    CNN（MNIST：2 Conv + 3 FC）, CNN（Fashion-MNIST：6 Conv + 1 FC）, ResNet18（CIFAR-10）
    AbstractFed-ADR 用 gray-box 编排攻击绕过多种聚合防御，再用 Hessian 一致性检测和窗口内恢复把准确率拉回。

    Fed-ADR 是一套面向联邦学习的攻击、检测与恢复框架：编排服务器改写合谋客户端的梯度，使参数服务器的现有防御把投毒更新当成良性，同时在发现后用历史梯度把全局模型原地修回来。

    Paper analysisRead analysis
  4. DefensesarXiv 2608.05578

    AMS 工具通过激活分析检测语言模型的安全训练改动

    提出 AMS，用激活分离与方向相似度检测安全训练改动

    arXiv
    2608.05578
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    Llama-3.1-8B-Instruct, Llama-3.2-3B-Instruct, Llama-3.2-1B-Instruct and others14
    AbstractAMS 按激活签名把安全改动分四类，前三类可由两级扫描覆盖，行为微调会漏。

    AMS 是面向开放权重模型、部署前使用的激活扫描器，用来看安全训练被拿掉或改写之后，激活空间里还留着什么几何痕迹。

    Paper analysisRead analysis
  5. DefensesarXiv 2608.02271

    Z-PEFT: Zero-shot backdoor detection for parameter-efficient fine-tuning based on canonical spectral features

    Propose Z-PEFT, which detects PEFT backdoors in a zero-shot manner using per-head spectral features.

    arXiv
    2608.02271
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Z-PEFT, Llama-2-7B, Llama-2-13B and others6
    Paper analysisRead analysis
  6. DefensesarXiv 2602.03085

    Triggers in the Haystack: Extracting and Reconstructing LLM Backdoor Triggers

    提出推理扫描器，在仅有模型文件时检测后门并重构触发词

    arXiv
    2602.03085
    Published
    Layer
    Model layer
    Setting
    Models and APIs
    Evaluated models
    gemma-3-270m-it, DeepSeek-R1-Distill-Qwen-1.5B, Phi-4-mini-instruct and others8
    Abstract用记忆泄漏和内部信号重构触发词。

    这是一个只做推理的 sleeper agent 扫描器：在不知触发词和目标行为时，检测因果语言模型里的条件后门并重构触发词。。问题在于，后门只在触发词出现时偏离基线，而开放仓库和网页投毒使检测方通常没有触发词、目标或激活语料。

    Paper analysisRead analysis
You have reached the end