Skip to content

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

Oct 13 · Tuesday

All papers

Found 6 papers
Filter5

Model risks

Deception and scheming18Reward hacking7Misalignment and value divergence3Dangerous agent actions13Hallucination and fabrication3

Defense type

Monitoring and auditing5

Impact

Unauthorized actions3Output manipulation3

System type

AI Agent4

Layer

Model layer5Application layer6

System component

Memory9Monitors5Reasoning chains2Permissions and sandboxing3
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 215papers have no tags yet and are not included in filtered results.
Model risks：Misalignment and value divergenceModel risks：Hallucination and fabricationLayer：Application layerSystem component：MonitorsSystem component：Reasoning chainsClear all

Another 215papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2610.07570

    ProbeGuard：按共识形成过程判断弃答，识别医疗问答中的一致错误

    提出 ProbeGuard，按共识形成过程对医疗问答做选择性弃答

    arXiv
    2610.07570
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Qwen3-8B, Llama-3.1-8B-Instruct, HuatuoGPT-o1-8B
    AbstractProbeGuard 用共识形成过程做可认证弃权，并在全票层分开答对与答错的共识。

    ProbeGuard 是给多轮共识医学问答用的认证弃权框架：基板照原样作答，弃权只看共识是怎么形成的。

    Paper analysisRead analysis
  2. DefensesarXiv 2609.15803

    研究者提出 k-robust 联盟对齐，用多智能体评审委托授权

    提出 k-robust 联盟对齐，刻画误对齐评审下的安全委托授权

    arXiv
    2609.15803
    Published
    Layer
    Application layer
    Setting
    Multi-agent systems
    Abstract联盟覆盖使无个体对齐的评审面板仍可相对基线安全授权，数值分数改善完备性。

    委托授权的安全不要求每位评审者对齐，而要求面板在效用空间覆盖委托人。

    Paper analysisRead analysis
  3. DefensesarXiv 2609.04159

    Sentinel-RL：将拓扑推理从 LLM 智能体中卸载到安全运营中心

    提出 Sentinel-RL，将拓扑处置从 LLM 卸载到受约束图策略

    arXiv
    2609.04159
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    HetGAT, PPO
    AbstractSentinel-RL 用图编码器与 PPO 承担拓扑决策，LLM 负责叙述与门控，并在 LANL 上报告检测与时延。

    Sentinel-RL 是面向 SOC 的神经符号架构：拓扑决策离开 LLM，改由认证图上的 HetGAT 与 PPO 完成。

    Paper analysisRead analysis
  4. DefensesarXiv 2608.27348

    INTENT-AS-A-TOOL：用意图工具追踪智能体失准

    提出 INTENT-AS-A-TOOL，用意图工具概率跟踪并干预智能体失准

    arXiv
    2608.27348
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Gemma-4-31B-IT, Qwen3.5-27B, Qwen3-235B-A22B and others5
    Abstract用意图工具的下一动作概率跟踪推理中的失准承诺，并在其成为首选时插入反思。

    Zhang 等人研究推理型语言模型在职场邮件沙箱里如何走向智能体失准，并提出用动作偏好而不是事后文本标签来跟踪承诺。问题是：有害执行往往出现在推理已经形成意图之后，但 CoT 监控是事后、粗粒度、依赖外部评审的，也看不到模型倾向调用哪个动作。

    Paper analysisRead analysis
  5. Evaluation and benchmarksarXiv 2608.00711

    SCHEMA：面向科学 Agent 幻觉的拓扑感知评测框架

    提出 SCHEMA，用概念图与轨迹分评测科学 Agent 幻觉

    arXiv
    2608.00711
    Published
    Layer
    Application layer
    Setting
    science agent
    Evaluated models
    Kimi-K2.5, Qwen3.6-Plus, Llama-4-Scout and others10
    AbstractSCHEMA 用知识拓扑评估科学智能体幻觉，终答正确仍可能来自有缺陷的中间推理。

    SCHEMA 是面向科学智能体幻觉的证据锚定评测框架，实例化在生物医学的 Protein Domain 与 PathVQA-Enhanced。作者在摘要中称它是首个同时做到证据锚定和拓扑感知的此类框架。

    Paper analysisRead analysis
You have reached the end