Skip to content

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

Oct 12 · Monday

All papers

Found 4 papers
Filter4

Defense type

Model hardening11Inference-time intervention4Detection and filtering3Instruction–data separation0Permissions and approval1Monitoring and auditing4

System type

Models and APIs3

System component

MCP and skills26Memory18RAG7Guardrails and judges69Monitors11Reasoning chains4Permissions and sandboxing63Vision14Video3Image generation4Multimodal20
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 265papers have no tags yet and are not included in filtered results.
Defense type：Detection and filteringDefense type：Instruction–data separationDefense type：Permissions and approvalSystem component：Reasoning chainsClear all

Another 265papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2609.36641

    从仅结果奖励的强化学习中诱导过程监督

    提出 TIPS，用仅结果奖励诱导生成式过程监督

    arXiv
    2609.36641
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Qwen3-4B-Thinking-2507, Qwen3-4B-Instruct-2507, Qwen3-8B-Thinking and others8
    Abstract结果监督的 GRPO 可诱导步骤验证。

    TIPS 用只含结果对错的强化学习训练生成式奖励模型，让步骤级验证出现在思维链里。

    Paper analysisRead analysis
  2. DefensesarXiv 2609.36562

    ThinkingGuard：通过逐步风险归因解码多模态大语言模型中的隐式危害

    构建 TriggerBench 并训练 ThinkingGuard，检测图文隐式风险

    arXiv
    2609.36562
    Published
    Layer
    Prompt layer
    Setting
    Models and APIs
    Evaluated models
    ThinkingGuard（Qwen3-VL-8B-Instruct）
    AbstractTriggerBench 隔离触发条件，ThinkingGuard 用分步归因在隐式风险检测上取得作者报告的较高平均分。

    这篇工作研究多模态隐式风险检测：图文各自良性，风险只在关键元素与触发元素交互时出现。

    Paper analysisRead analysis
  3. DefensesarXiv 2609.35659

    Tracekit：面向自主编码 Agent 的防篡改意图-推理-动作审计系统

    提出 Tracekit，审计编码 Agent 的意图、推理与动作并检测账本篡改

    arXiv
    2609.35659
    Published
    Layer
    Application layer
    Setting
    coding agent
    Evaluated models
    Claude Code (claude -p, version 2.1), 独立评审（同一 CLI，无工具）
    AbstractTracekit 把意图、自述和动作写入可锚定的哈希链。

    Tracekit 是无第三方依赖的编码智能体审计层：把人类请求、模型自述和已执行动作并排写入可外部锚定的哈希链，执行前用正则门，事后用规则和独立模型互查。

    Paper analysisRead analysis
  4. DefensesarXiv 2608.12876

    SPARED：用对抗编辑数据训练基于推理的 AI 生成图像检测器

    提出 SPARED，用对抗编辑数据训练基于推理的 AI 生成图像检测器

    arXiv
    2608.12876
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Attacker
    Black-box, White-box
    Evaluated models
    Qwen3.5-9B
    AbstractSPARED 用防捷径的编辑对抗环训练 9B 推理检测器。

    SPARED 是一套用于可解释 AI 生成图像检测的攻防强化学习流程，报告模型是 Qwen3.5-9B 上的第三轮防御器。

    Paper analysisRead analysis
You have reached the end