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

Oct 12 · Monday

All papers

Found 51 papers

Filter3

Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 261papers have no tags yet and are not included in filtered results.

System type：web agent System type：Embodied systems and robotics Layer：Application layer Clear all

Another 261papers do not have category tags yet and are not included in filtered results.

OtherarXiv 2610.13479Oct 9
A local agentic AI framework generates edge-case scenarios for safety assurance
Proposes a local dual-model closed loop to generate edge scenarios for autonomous system safety assurance
arXiv
2610.13479
Published
Oct 9
Layer
Application layer
Setting
LLM applications, Embodied systems and robotics
Evaluated models
PPO
AbstractLocal closed-loop generation of JSON edge cases.
The authors built a local agentic loop that runs on a laptop to generate edge cases for safety arguments of autonomous systems, rather than releasing a new public test set.
Paper analysisRead analysis
OtherarXiv 2610.13376Oct 9
EndoRSI: Recursive Self-Improvement of Endoscopic Robot Policies Driven by Physical Experiment Evidence
Proposes EndoRSI, recursively improving endoscopic robot policies with physical evidence
arXiv
2610.13376
Published
Oct 9
Layer
Application layer
Setting
Embodied systems and robotics
Evaluated models
Qwen3-VL-4B
Paper analysisRead analysis
OtherarXiv 2610.13372Oct 9
Toward Safe World Models: Risks, Defenses, and Evaluation
A survey of risks, defenses, and evaluation in the six-stage lifecycle of world models
arXiv
2610.13372
Published
Oct 9
Layer
Application layer
Setting
Embodied systems and robotics
AbstractSafety is written as a property of the prediction-decision closed loop; claims must be linked to observed consequences and versions.
World model safety is organized by the nature of the prediction-decision loop, not by isolated prediction scores.
Paper analysisRead analysis
OtherarXiv 2610.12369Oct 9
Embodied Turing Machine: Code-Only-as-Policy for Recursive Self-Improvement in Robots with Stateful Code
Propose COAP, which uses stateful code to control a robot at test time without a model
arXiv
2610.12369
Published
Oct 9
Layer
Application layer
Setting
Embodied systems and robotics
Paper analysisRead analysis
DefensesarXiv 2610.09469Oct 7
Secure-CUA: Constraining the Impact of Untrusted Content in Computer-Use Agents
提出 Secure-CUA，用动作事务约束计算机使用智能体的不可信影响
arXiv
2610.09469
Published
Oct 7
Layer
Application layer
Setting
web agent
Evaluated models
Claude Opus 5, Gemini 3.8 Flash, GPT-5.6-Sol
- DefensesInstruction–data separation
- AttacksPrompt injection
- ComponentVision
AbstractSecure-CUA 用每步事务约束不可信影响，良性 TSR 接近 Vanilla-CUA。
Secure-CUA 要限制图形界面里不可信内容对 CUA 决策和 GUI 执行的影响，同时保留合法任务所需的不可信信息。
Paper analysisRead analysis
AttacksarXiv 2610.09240Oct 7
WebMirage hijacks visual web agents' browser operations with adversarial images
提出 WebMirage，用局部对抗图像劫持视觉网页 Agent 的浏览器操作
arXiv
2610.09240
Published
Oct 7
Layer
Application layer
Setting
web agent
Evaluated models
LLaVA-v1.5-13B, LLaVA-v1.6-34B, MiniCPM-o-8B and others6
- AttacksPrompt injection
- TechniqueGradient and representation optimization
- ComponentVision
Abstract论文提出了针对网络智能体的端到端视觉红队框架 WebMirage，揭示了从视觉定位到浏览器执行的安全脆弱性。
WebMirage 是一个面向视觉定位网络智能体的流水线感知端到端红队攻击框架。
Paper analysisRead analysis
DefensesarXiv 2610.08773Oct 7
AdvSim2Real 用任务与注入对手共同演化训练网页智能体
提出 AdvSim2Real，协同训练网页智能体抵御页面注入
arXiv
2610.08773
Published
Oct 7
Layer
Application layer
Setting
web agent
Evaluated models
Qwen3.5-4B
- DefensesModel hardening
- AttacksPrompt injection
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.08871Oct 6
CredLeak-Bench
提出 CredLeak-Bench，评测 Agent 面对钓鱼的凭据泄露与恢复
arXiv
2610.08871
Published
Oct 6
Layer
Application layer
Setting
web agent
Evaluated models
Claude Sonnet 5, Claude Opus 4.8, Gemini 3.6 Flash and others7
- AttacksPrompt injection
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.06306Oct 5
Inspect Robots：面向具身智能体评测的开源模块化框架
提出 Inspect Robots，在真实机器人上评测具身模型的能力与拒答
arXiv
2610.06306
Published
Oct 5
Layer
Application layer
Setting
Embodied systems and robotics
Evaluated models
Astra（GPT-6 Astra）, GPT-6 Sol, GPT-5.6 Sol and others7
- RiskInappropriate refusal
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.05586Oct 5
AgentDoxx：联网搜索 Agent 对匿名访谈文本的再识别评测
构建 AGENTDOXX 基准，评测联网搜索 Agent 对匿名访谈的再识别
arXiv
2610.05586
Published
Oct 5
Layer
Application layer
Setting
web agent
Evaluated models
GPT-5.6 Terra, GPT-5 Mini, GPT-4.1 Mini and others10
- AttacksMalicious use
- ComponentRAG
Abstract论文提出了已知身份的重识别基准 AGENTDOXX，揭示搜索与记忆双重风险及现有防御在检索轨迹中的泄露漏洞。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.04366Oct 3
CrashSim 用真实事故先验生成碰撞场景并构建 nuCrash 数据集
提出 CrashSim 用真实事故先验引导扩散仿真并构建 nuCrash 评测自动驾驶规划器
arXiv
2610.04366
Published
Oct 3
Layer
Application layer
Setting
Embodied systems and robotics
Evaluated models
Lane-graph planner, PDM-closed planner, IL planner and others5
AbstractCrashSim 迁入真实事故先验生成场景。
CrashSim 用检索到的真实事故先验引导扩散式多车仿真，为自动驾驶安全评测生成碰撞与近碰撞场景。。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.03574Oct 3
HyperBrowseComp：面向网页浏览 Agent 的多语言多模态压力测试基准
构建网页浏览 Agent 的多语言多模态压力测试基准 HyperBrowseComp
arXiv
2610.03574
Published
Oct 3
Layer
Application layer
Setting
web agent
Evaluated models
Gemini 3.1 Pro Preview, Gemini 3.7 Flash, GPT-5.6 Terra and others5
Paper analysisRead analysis
OtherarXiv 2610.03036Oct 2
WebFovea 提出视觉网页智能体四阶段加固框架，WebRetriever 挑战赛得分从 31.0 升至 57.0
提出 WebFovea，加固视觉网页智能体的解析、执行、反馈与观察往返
arXiv
2610.03036
Published
Oct 2
Layer
Application layer
Setting
web agent
Evaluated models
claude-opus-4-6
- ComponentVision
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.02874Oct 2
SceneFactory-3D：将 2D 交通场景提升为 3D 物理反事实以做安全评测
提出 SceneFactory-3D，用路况反事实评测驾驶策略的闭环安全性
arXiv
2610.02874
Published
Oct 2
Layer
Application layer
Setting
Embodied systems and robotics, Multi-agent systems
Evaluated models
Frozen L0, L0-T, L1 and others5
AbstractSceneFactory-3D 固定场景只改路况。
SceneFactory-3D 是 GPU 批处理的多智能体驾驶仿真器，用来在场景和控制器固定时只改路面物理，并比较闭环结果。
Paper analysisRead analysis
OtherarXiv 2610.02662Oct 2
论文发现语言驱动机器人安全监控的细化缺口：安全高层计划可能产生不安全执行
审计机器人安全监控的细化缺口，比较表面计划与图细化轨迹的裁决
arXiv
2610.02662
Published
Oct 2
Layer
Application layer
Setting
Embodied systems and robotics
- ComponentMonitors
Abstract表面轨迹可漏掉图诱导的禁区进入。
作者审计 RoboGuard 的一个假设：对高层动作轨迹的安全裁决，在补上语义图诱导的导航和已记录隐式效应之后是否仍然成立。做法是不改验证器，并固定同一条 LTL。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2609.38948Sep 30
DrivingBench 发布：让视觉语言模型驾驶丰田 Corolla 绕桩
提出 DrivingBench，评测通用视觉语言模型能否闭环驾驶真车
arXiv
2609.38948
Published
Sep 30
Layer
Application layer
Setting
Embodied systems and robotics
Evaluated models
GPT-6 Astra, Claude Fable 5.1, Grok 4.6 and others4
- ComponentVision
Abstract仅 Astra 第二次跑完低速锥桶路线。
DrivingBench 评测通用 VLM 能否自己闭环开真车，推理延迟也算进任务。
Paper analysisRead analysis
DefensesarXiv 2609.34974Sep 28
Veer：网页 Agent 运行时防御暗模式
提出 Veer，在网页 Agent 执行前改写会造成未授权后果的网页状态
arXiv
2609.34974
Published
Sep 28
Layer
Application layer
Setting
web agent
Evaluated models
GPT-5.6-Luna, DeepSeek-V4-Flash
- AttacksPrompt injection
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2609.32915Sep 27
AgentTell：浏览器 Agent 的行为侧信道泄漏基准
构建 AGENTTELL 基准，测量浏览器 Agent 的行为侧信道隐私泄露
arXiv
2609.32915
Published
Sep 27
Layer
Application layer
Setting
web agent
Evaluated models
Claude Sonnet 5, Gemini 3.7 Flash, GPT-5.6 Luna and others6
- RiskDangerous agent actions
Abstract论文评估了浏览器智能体的行为侧信道泄露风险，并基于近万次会话提供了评测数据。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2609.27273Sep 23
微软研究院发布 CAVEAT
提出 CAVEAT 基准，评测激励失配市场中计算机使用 Agent 的购买决策退化
arXiv
2609.27273
Published
Sep 23
Layer
Application layer
Setting
web agent
Evaluated models
GPT-5.6-Sol, GPT-5.6-Terra, GPT-5.6-Luna and others14
- RiskDangerous agent actions
Abstract提出 CAVEAT 基准揭示智能体在商业偏好下的决策退化，诊断出三类失效并给出缓解方案。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2609.26618Sep 22
NavSafe-∞：在真实感环境中评测闭环驾驶安全
提出 NAVSAFE-∞ 写真闭环基准，评测端到端驾驶策略的闭环安全
arXiv
2609.26618
Published
Sep 22
Layer
Application layer
Setting
Embodied systems and robotics
Evaluated models
LTF, DiffusionDrive, DrivoR and others13
- RiskDangerous agent actions
Abstract写真事件级闭环基准显示开环高分不能稳定变成闭环安全。
NAVSAFE-∞ 用写真、事件级闭环仿真检查端到端驾驶策略的安全，而不是只看开环轨迹分数。
Paper analysisRead analysis

AI Safety Daily

Read the latest issueOne issue per day, updated as events develop

All papers

Attack type

Attack technique

Model risks

Defense type

Impact

System type

Attack surface / entry point

Layer

System component

A local agentic AI framework generates edge-case scenarios for safety assurance

EndoRSI: Recursive Self-Improvement of Endoscopic Robot Policies Driven by Physical Experiment Evidence

Toward Safe World Models: Risks, Defenses, and Evaluation

Embodied Turing Machine: Code-Only-as-Policy for Recursive Self-Improvement in Robots with Stateful Code

Secure-CUA: Constraining the Impact of Untrusted Content in Computer-Use Agents

WebMirage hijacks visual web agents' browser operations with adversarial images

AdvSim2Real 用任务与注入对手共同演化训练网页智能体

CredLeak-Bench

Inspect Robots：面向具身智能体评测的开源模块化框架

AgentDoxx：联网搜索 Agent 对匿名访谈文本的再识别评测

CrashSim 用真实事故先验生成碰撞场景并构建 nuCrash 数据集

HyperBrowseComp：面向网页浏览 Agent 的多语言多模态压力测试基准

WebFovea 提出视觉网页智能体四阶段加固框架，WebRetriever 挑战赛得分从 31.0 升至 57.0

SceneFactory-3D：将 2D 交通场景提升为 3D 物理反事实以做安全评测

论文发现语言驱动机器人安全监控的细化缺口：安全高层计划可能产生不安全执行

DrivingBench 发布：让视觉语言模型驾驶丰田 Corolla 绕桩

Veer：网页 Agent 运行时防御暗模式

AgentTell：浏览器 Agent 的行为侧信道泄漏基准

微软研究院发布 CAVEAT

NavSafe-∞：在真实感环境中评测闭环驾驶安全