Translation in progress Some entries still include text in the original language.
Oct 12 · Monday
All papers Found 37 papers Another 265 papers do not have category tags yet and are not included in filtered results.
“
About this collection Collection in progress We are building a large AI safety paper collection to help researchers find work related to their interests. Papers are collected automatically andfiltered only for relevance to AI safety , without judging quality, novelty or reading value. The default “Annotated first” shows papers with tags or published paper analyses first. Each group is ordered by original publication date, newest first, with undated papers last. “Newest first” uses publication dates only. This order is not a quality ranking.
One-line summaries, tags and paper analyses are prepared by AI from the papers. Language models may have systematic biases and produce hallucinations.We plan to carry out human annotation and review to reduce errors in this content.
提出 Red-TTT，在攻击过程中更新攻击者实现自动化越狱
Published Oct 4
Evaluated models Llama-3-8B-Instruct , gpt-oss-20b , gpt-4.1-mini and others4 提出 CodeMimicry，以结构化代码补全诱导模型生成有害内容
Published Sep 30
Evaluated models GPT-4o , GPT-4.1 , GPT-5-chat and others10 提出 SceneJail，利用视频情境上下文越狱视频多模态模型
Published Sep 30
Evaluated models Qwen2.5-VL-7B-Instruct , Qwen2.5-VL-32B-Instruct , Qwen3-VL-8B-Instruct and others8 提出 FinRT，将消费金融自适应红队轨迹蒸馏为可复用对抗提示生成器
Published Sep 29
Evaluated models Llama-3.3-70B , Llama-3.1-8B , Mixtral-8x7B and others6 Abstract FinRT 用前向失效发现加校准代理 DPO，把消费金融红队收成单次前向生成器。FinRT 是面向消费金融的红队框架：先发现政策失效，再训练一个按政策、领域与目标行为出提示的生成器。
提出 RISE，迭代演化可复用策略对文生图模型做红队
Published Sep 28
Evaluated models DALL·E 3 , Nano Banana 2 , GPT-Image-2 and others7 消融比较方言越狱中表层形式、文化框架与策略库的作用
Published Sep 15
Evaluated models GPT-4o , DeepSeek-Reasoner , DeepSeek-Chat 提出 AdvPIE，在黑盒下搜索表面无害却生成有害图像的提示
Published Sep 5
Evaluated models Stable Diffusion 3.5 Large , FLUX.1-schnell , Safe-StableDiffusion 1.5 and others5 提出 CRACK，用多智能体辩论在黑盒下越狱文生图多层安全过滤
Published Sep 1
Evaluated models Stable Diffusion v1.4 , SDXL , DALL·E 3 and others4 Abstract CRACK 以分层诊断的多智能体辩论搜索多层 T2I 过滤器的联合规避区域。CRACK 是面向文本生成图像模型多层安全栈的黑盒越狱框架，用几何上的 Detection Surface 解释为何粗粒度反馈不够。
提出 RedEvoAgent，演化攻击技能对黑盒智能体做自动红队
Published Aug 28
Evaluated models MiniMax-M2.5 , DeepSeek-V4-Flash , Qwen3.5-35B Abstract RedEvoAgent 用验证棘轮演化可读攻击技能，在两种产品级 harness 上多数设置高于固定工具与红队基线。RedEvoAgent 是面向产品级 black-box 智能体的自动红队方法，用一份可演化、人类可读的攻击技能编排多个越狱工具。 。
提出 MAMJ，在元层面交替优化多模态越狱的策略提示与攻击者权重
Published Aug 27
Evaluated models GPT-4o , Gemini-3-Pro-Preview , Seed 2.0 提出 JailbreakSkill，用可演化技能自动搜索并复用越狱提示
Published Aug 17
Evaluated models GLM 5.2 , DeepSeek-V4-Pro Preview , Llama-3.3-70B-Instruct and others8 提出 ARENA 音频红队框架，自动构造文本安全且音频接地的越狱输入
Published Aug 16
Evaluated models Audio Flamingo 3 , Qwen2-Audio-7B-Instruct , MiMo-Audio-7B-Instruct and others4 训练自博弈红队智能体 GPT-Red，规模化生成提示注入与越狱攻击
Published Jul 29
Evaluated models GPT-5.6 , GPT-5.5 , GPT-5.4 and others8 提出 AHA 自动科研循环，发现生产 Agent 的可复用漏洞概念
Published Jul 13
Evaluated models Minimax-M2.7 , Kimi-K2.6 , Deepseek-V4-Pro and others5 Abstract 论文提出基于可证伪科研循环的 AHA 框架，从生产级智能体中挖掘并沉淀漏洞概念图，提升攻击复用率并指导防御加固。AHA 是一项针对生产级 LLM 智能体自动化红队测试的研究，旨在解决传统方法仅保留具体攻击载荷而缺乏成因解释、导致跨版本复用困难且无法指导修复的问题。
提出 INFER+ 信息过载方法，越狱大型视觉语言模型
Published Jul 3
Evaluated models Qwen3-VL-8B , Qwen2-VL-7B , InternVL3.5-8B and others8 Abstract 用图文过载越狱 LVLM。这篇工作用信息过载 越狱 LVLM，而不是靠 OOD 变换隐藏意图。
提出 UniAttack 单轮黑盒越狱框架，融合攻击特征绕过多层防御
Published Jun 15
Evaluated models DeepSeek-Chat , DeepSeek-Reasoner , GPT-4.1-Nano and others9 提出 JailbreakOPT，组合原子工具迭代优化黑盒单轮越狱提示
Published Jun 10
Evaluated models Vicuna-13b-v1.5 , LLaMA-3-8b , GPT-4o and others8 提出 MemoAttack，用技能结构化记忆做黑盒越狱
Published May 28
Evaluated models kimi-k2.5 , minimax-M2.1 , qwen3.5-397B-A17B 提出 Babel 混淆分布采样，黑盒越狱闭源 LLM 安全注意力
Published May 18
Evaluated models gpt-4o-2024-11-20 , Gemini-2.0-flash , Claude-3-5-haiku-20241120 and others4 提出 Stable-GFlowNet，用对比轨迹平衡训练红队策略生成多样高毒性攻击
Published May 1
Evaluated models Qwen2.5-1.5B-Instruct , Gemma3-4B-Instruct , Llama3.2-3B-Instruct and others5 Abstract S-GFN 用 CTB、NGP 和 MKS 训练红队攻击者。S-GFN 是用于 LLM 红队的 GFlowNet 训练方法，用轨迹之间的相对比较代替配分函数 Z。 。