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

Oct 12 · Monday

All papers

Found 248 papers

Filter5

System component

MCP and skills26 Memory11 Guardrails and judges93 Monitors48 Reasoning chains64 Permissions and sandboxing12 Fine-tuning and open weights60 Vision101 Audio32 Video11 Image generation46 Embeddings4 Multimodal180

Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 233papers have no tags yet and are not included in filtered results.

System type：Models and APIs System type：coding agent System component：Guardrails and judges System component：Video System component：Multimodal Clear all

Another 233papers do not have category tags yet and are not included in filtered results.

DefensesarXiv 2610.13733Oct 9
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
Oct 9
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
- AttacksMalicious use
- ComponentImage generation
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
DefensesarXiv 2610.13356Oct 9
SWE-Critix: Post-trained general-purpose verifier for evaluating coding agent trajectories
Proposing SWE-Critix, a post-trained general verifier for evaluating coding agent trajectories
arXiv
2610.13356
Published
Oct 9
Layer
Application layer
Setting
coding agent
Evaluated models
Qwen3-30B-A3B-Thinking-2507, SWE-Critix SFT, SWE-Critix RL
- DefensesDetection and filtering
- ComponentGuardrails and judges
Paper analysisRead analysis
DefensesarXiv 2610.13333Oct 9
LETHE: Defending Against Unauthorized Image-to-Video Generation via Value Diversity Collapse
Proposes LETHE, which perturbs reference images to block unauthorized image-to-video
arXiv
2610.13333
Published
Oct 9
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
Wan2.1-I2V-14B, CogVideoX-5B-I2V, LightX2V and others6
- AttacksMalicious use
- ComponentImage generation
AbstractLETHE suppresses the diversity of reference values and attracts attention, making protected images difficult to animate faithfully.
LETHE is a protective perturbation added to personal images before publication, used to counter unauthorized image-to-video generation.
Paper analysisRead analysis
Analysis and theoryarXiv 2610.12083Oct 8
Paper audits LLM judge reliability: repeated agreement does not equal correctness
Auditing LLM judge reliability and defining trustworthy verdict rate and accuracy upper bound
arXiv
2610.12083
Published
Oct 8
Layer
Model layer
Setting
Models and APIs
Evaluated models
GPT-5.2, GPT-4o, Gemini-2.5-Pro and others6
- ComponentGuardrails and judges
AbstractThe authors claim that comparing judges should look at T, not just accuracy.
The authors audit whether LLM-as-a-Judge decisions can still be treated as actionable ground truth under repeated queries, swapped candidate order, and changed prompt formats.
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.11773Oct 8
研究提出 Agent 安全新维度：识别未履行的安全义务
提出义务识别基准 ObligationBench，并训练护栏识别 Agent 未履行的安全义务
arXiv
2610.11773
Published
Oct 8
Layer
Application layer
Setting
coding agent
Evaluated models
ObligationGuard, Claude-Opus-4.8, GPT-5.6-Sol and others17
- DefensesDetection and filtering
- RiskDangerous agent actions
- ComponentGuardrails and judges
Abstract义务是护栏缺口。
作者认为编码智能体的安全不能只靠拦截禁止动作，还要找出终止时仍未做的安全关键动作，并将其称为义务。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.11599Oct 8
研究：大语言模型版本更替削弱 AI 辅助科研写作筛查
测量 LLM 版本更替对科学摘要 AI 改写检测器的影响
arXiv
2610.11599
Published
Oct 8
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
RoBERTa-base, Pangram
- DefensesDetection and filtering
- ComponentGuardrails and judges
Abstract作者称检测器的基准准确率只是临时的，每次发布都要连同更早版本重验。
作者量化的是：科学摘要的 AI 改写筛查，在 LLM 版本不断更换、而人类摘要误标必须少见时，会怎样变。问题来自期刊和会议已经按分数阈值筛查投稿，但基准多停在评估时可得的版本上。
Paper analysisRead analysis
DefensesarXiv 2610.11281Oct 8
包络采样：用少量真值标注重校准 LLM 评委以缓解奖励作弊
提出包络采样，用少量真值标注重标定 LLM 评委以缓解奖励作弊
arXiv
2610.11281
Published
Oct 8
Layer
Training and data layer
Setting
Models and APIs
Evaluated models
Qwen3-4B, Qwen3.8-27B, Qwen2.5-14B-Instruct
- DefensesModel hardening
- RiskReward hacking
- ComponentGuardrails and judges
Abstract包络采样把真值标注打到代理尾部，同样预算下缓解 reward hacking。
作者研究后训练前如何选择真值标注的查询分布，使重标定后的廉价代理更适合被优化。真奖励昂贵时，实践用 LLM judge 代替人类。若失准只发生在基座很少生成的输出上，on-policy 标注修不到这些点，随后的 KL 正则优化会把质量集中到被高估的输出。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.11112Oct 8
EpiReal-Bench：商用图像生成模型虚假声明红队基准
构建 EpiReal-Bench，评测图像生成器将虚假主张渲染为可信证据
arXiv
2610.11112
Published
Oct 8
Layer
Prompt layer
Setting
Models and APIs
Attacker
Black-box
Evaluated models
GPT-Image-2, Nano Banana 2, Grok Imagen 2.0 and others4
- AttacksJailbreaks
- ComponentImage generation
Abstract四款商用图像模型能把假主张画成可信证据。
EpiReal-Bench 用来量商用图像生成器会不会把虚假现实主张画成可信视觉证据。
Paper analysisRead analysis
InterpretabilityarXiv 2610.10865Oct 8
用路由稀疏自编码器解耦语音编码器中的语言学与副语言学信息
提出路由 TopK SAE，把语音编码器稀疏码拆成语言与副语言通路
arXiv
2610.10865
Published
Oct 8
Layer
Model layer
Setting
Models and APIs
Evaluated models
SPEAR XLarge, WavLM Large
- ComponentAudio
Abstract路由 SAE 在冻结 SPEAR 与 WavLM 上分开语言和副语言因子，探针与交换支持通路分工。
TopK SAE 的稀疏码被分成语言通路与副语言通路，用来分开冻结语音编码器里同时存在的音素信息和说话人、情感、韵律。只重构不能得到作者选定的这一分解。语言通路由音素识别监督，副语言通路由说话人识别、情感分类和韵律预测监督，作用在对方通路上的梯度反转对抗用来压低交叉因子。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.10358Oct 8
Open-MMUnlearning 发布：统一多模态大模型遗忘方法与评测框架
提出统一框架评测多模态大模型的遗忘方法
arXiv
2610.10358
Published
Oct 8
Layer
Training and data layer
Setting
Models and APIs
Evaluated models
LLaVA-1.5-7B
- DefensesModel hardening
- ComponentVision
Paper analysisRead analysis
DefensesarXiv 2610.09941Oct 7
OrthoPurify 用一步正交投影净化被后门污染的视觉语言模型
提出 OrthoPurify，用正交投影清除视觉语言模型适配器中的劫持方向
arXiv
2610.09941
Published
Oct 7
Layer
Training and data layer
Setting
Models and APIs
Evaluated models
LLaVA-1.5-7B, Qwen3-VL-8B, LLaVA-1.5-13B and others5
- DefensesModel hardening
- AttacksPoisoning and backdoors
- ComponentFine-tuning and open weights
Abstract一次投影去掉劫持方向。
OrthoPurify 是部署前的 LVLM 后门净化方法，用一次正交投影去掉被劫持的权重更新方向。
Paper analysisRead analysis
AttacksarXiv 2610.09819Oct 7
FAB 攻击：给 HuBERT/WavLM 植入后门，背景警笛声即可让下游语音任务失灵
提出 FAB 攻击，向声学基础模型植入可由环境声触发的后门
arXiv
2610.09819
Published
Oct 7
Layer
Training and data layer
Setting
Models and APIs
Evaluated models
HuBERT-base, WavLM-base
- AttacksPoisoning and backdoors
- TechniqueDormant triggers
- ComponentAudio
Abstract论文提出了针对声学基模的 FAB 后门攻击，在弱威胁模型下实现了任务无关、物理可触发的后门植入。
Paper analysisRead analysis
DefensesarXiv 2610.09703Oct 7
研究揭示 Token-Pruning 对 VLM 安全性的影响并提出 SAP 缓解机制
提出 Safety-Aware Pruning，缓解视觉 token 剪枝放大的多模态越狱
arXiv
2610.09703
Published
Oct 7
Layer
Model layer
Setting
Models and APIs
Attacker
Black-box
Evaluated models
LLaVA-1.5-7B
- DefensesInference-time intervention
- AttacksJailbreaks
- ComponentVision
Abstract剪枝去掉背景 token 后注意力塌向恶意前景。
这篇工作评估视觉 token 剪枝如何改变 VLM 在多模态越狱下的行为，并给出推理时防御 SAP。
Paper analysisRead analysis
AttacksarXiv 2610.10612Oct 7
PyCache Trap 利用 Python 缓存替换绕过七款 Agent Skill 扫描器
提出 PyCache Trap，用字节码缓存替换绕过 Agent 技能扫描器
arXiv
2610.10612
Published
Oct 7
Layer
Application layer
Setting
coding agent
Evaluated models
CISCO, HSS, AUDITOR and others8
- DefensesDetection and filtering
- AttacksDetection evasion
- TechniqueDefense-targeted
- ComponentMCP and skills
Abstract论文剖析智能体技能审查与执行脱节盲区，提出 PyCache Trap 攻击并构建基于可信复现的 EAV 防御。
这篇论文研究了智能体技能（Agent Skills）安全扫描中的“审查–执行脱节”问题，并提出了对应的攻击验证与防御方案。
Paper analysisRead analysis
DefensesarXiv 2610.09167Oct 7
Persist-3 用顺序检验识别未见越狱来源，提升 MTK 与 Prompt Guard 2 防御效果
提出 Persist-3 序贯检验，从隐藏表示识别未见越狱源
arXiv
2610.09167
Published
Oct 7
Layer
Model layer
Setting
Models and APIs
Evaluated models
Llama2-7B-chat, Qwen2.5-7B-Instruct, Gemma-2-9B-it and others4
- DefensesMonitoring and auditing
- AttacksJailbreaks
- ComponentMonitors
AbstractPersist-3 用有限窗口检验攻击流是否超出已知越狱源，检出率低于全历史 SR-DKW，但决策只依赖最近观测。
Persist-3 是一个有限记忆的序贯检测器，用来判断攻击表示流是否已经不能由任何一个已知越狱源解释。。
Paper analysisRead analysis
AttacksarXiv 2610.09027Oct 7
P-VMI：对抗图像经 KV cache 持续影响多轮对话
提出持久视觉记忆注入，使对抗图像经缓存在触发时操控输出
arXiv
2610.09027
Published
Oct 7
Layer
Infrastructure layer
Setting
Models and APIs
Evaluated models
Qwen3-8B, Qwen3-VL-8B-Instruct, Gemma-4-12B
- AttacksPrompt injection
- TechniqueDormant triggers
- ComponentVision
Abstract论文提出 P-VMI，证实对抗输入被掩蔽后仍可经 KV 缓存持久控制模型，强调了保留状态的安全风险。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.08791Oct 7
研究者发布视频世界模型物理一致性基准 World Models' Last Exam in Physics
提出 World Models' Last Exam 评测视频物理一致性
arXiv
2610.08791
Published
Oct 7
Layer
Model layer
Setting
Models and APIs
Evaluated models
Seedance-2.5, MiniMax-H3, Cosmos3-Super and others8
- ComponentImage generation
Paper analysisRead analysis
InterpretabilityarXiv 2610.08887Oct 6
Moshi 全双工语音模型的情感引导遵循几何结构而非标签
用均值差激活转向检验全双工语音模型的情感方向几何
arXiv
2610.08887
Published
Oct 6
Layer
Model layer
Setting
Models and APIs
Evaluated models
kyutai/moshika-pytorch-q8
- ComponentAudio
Abstract转向跟方向几何走而不跟标签走：sad 可单独对准，高唤醒情感共享方向。
Moshi 上的情感激活转向跟方向之间的几何关系走，不跟被请求的标签走。
Paper analysisRead analysis
AttacksarXiv 2610.08331Oct 6
STCA：研究者提出针对自动驾驶视频 VLM 的时空一致性对抗攻击
提出 STCA，对驾驶视频 VLM 构造可迁移的时空对抗扰动
arXiv
2610.08331
Published
Oct 6
Layer
Prompt layer
Setting
Models and APIs
Attacker
No-box (offline transfer)
Evaluated models
Video-LLaVA-7B, Qwen2.5-VL-7B, Dolphins
- AttacksDetection evasion
- TechniqueGradient and representation optimization
- ComponentVideo
Paper analysisRead analysis
Analysis and theoryarXiv 2610.08178Oct 6
研究给出潜空间图像水印的鲁棒性上限理论分析
分析潜空间图像水印对像素扰动的内在鲁棒性上限
arXiv
2610.08178
Published
Oct 6
Layer
Model layer
Setting
Models and APIs
Evaluated models
Stable Diffusion v2-1
- DefensesWatermarking and provenance
- ComponentImage generation
Abstract作者称当前潜空间水印在低 FPR 下受扩张映射与分布偏移的限制。
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

Attacker access

Attack surface / entry point

Layer

System component

KVeil: First-Place Solution for the ECCV 2026 LifeGenIP Unlearnable Video Challenge

SWE-Critix: Post-trained general-purpose verifier for evaluating coding agent trajectories

LETHE: Defending Against Unauthorized Image-to-Video Generation via Value Diversity Collapse

Paper audits LLM judge reliability: repeated agreement does not equal correctness

研究提出 Agent 安全新维度：识别未履行的安全义务

研究：大语言模型版本更替削弱 AI 辅助科研写作筛查

包络采样：用少量真值标注重校准 LLM 评委以缓解奖励作弊

EpiReal-Bench：商用图像生成模型虚假声明红队基准

用路由稀疏自编码器解耦语音编码器中的语言学与副语言学信息

Open-MMUnlearning 发布：统一多模态大模型遗忘方法与评测框架

OrthoPurify 用一步正交投影净化被后门污染的视觉语言模型

FAB 攻击：给 HuBERT/WavLM 植入后门，背景警笛声即可让下游语音任务失灵

研究揭示 Token-Pruning 对 VLM 安全性的影响并提出 SAP 缓解机制

PyCache Trap 利用 Python 缓存替换绕过七款 Agent Skill 扫描器

Persist-3 用顺序检验识别未见越狱来源，提升 MTK 与 Prompt Guard 2 防御效果

P-VMI：对抗图像经 KV cache 持续影响多轮对话

研究者发布视频世界模型物理一致性基准 World Models' Last Exam in Physics

Moshi 全双工语音模型的情感引导遵循几何结构而非标签

STCA：研究者提出针对自动驾驶视频 VLM 的时空一致性对抗攻击

研究给出潜空间图像水印的鲁棒性上限理论分析