Translation in progress Some entries still include text in the original language.
Oct 12 · Monday
All papers Found 48 papers Another 261 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.
Proposes digital lock compilation, encrypting executable networks into host weights to hide arbitrary capabilities
Published Oct 10
Evaluated models Gemma 3 270M Abstract The authors use a digital lock to hide an executable capability and state that elicitation on the chemistry and parity tasks did not recover the capability.The authors construct a class of backdoors that encrypt an executable network into the weights, used to discuss the limits of white-box elicitation audits under supply chain attacks.
Trace attention to identify harmfulness to the MLP refusal write pathway and propose Sentinel
Published Oct 9
Evaluated models Llama-3.1-8B-Instruct , Qwen2.5-7B-Instruct , Qwen3-8B-Instruct and others18 Abstract Attention updates are more readable for harmfulness, MLP updates write refusal, and Sentinel routes alerts using two-path signals.The authors reframe jailbreaking and excessive refusal of aligned models as a pathway problem on the residual stream from recognition to writing, and accordingly provide conditional reminders before decoding.
提出 PatchBench-Local 协议，评测激活补丁修复越狱的局部附带损伤
Published Oct 7
Evaluated models Qwen2.5-3B-Instruct , Qwen2.5-14B-Instruct , gemma-3-4b-it and others8 Abstract 论文提出 PatchBench 与局部评估协议，揭示了激活补丁在全局能力无损下仍存局部良性退化。
提出 Safety-Aware Pruning，缓解视觉 token 剪枝放大的多模态越狱
Published Oct 7
Evaluated models LLaVA-1.5-7B 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。
提出 Persist-3 序贯检验，从隐藏表示识别未见越狱源
Published Oct 7
Evaluated models Llama2-7B-chat , Qwen2.5-7B-Instruct , Gemma-2-9B-it and others4 Abstract Persist-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。 。
测量有损投机解码的安全代价，并提出 SecureSD 收紧早期验证
Published Oct 7
Evaluated models Qwen3-8B , Qwen3-32B , Llama-3.1-8B-Instruct and others6 Abstract 有损推测解码的早期偏离更快抬高越狱与注入风险。作者研究有损推测解码在换取速度时，是否会比效用更快地损害越狱与提示注入防护，并给出按位置校正的 SecureSD。
提出 PEV 攻击，在输入嵌入上加高斯噪声越狱开源模型
Published Oct 6
Evaluated models Llama-3.1-8B-Instruct , Qwen2.5-1.5B-Instruct , Phi-4-mini-3.8B and others6 Abstract PEV 用 embedding 高斯噪声，在六个模型上越狱 JailbreakBench 全部提示。PEV 是对 open-weight LLM 的 embedding 噪声越狱。
提出 ReSA，从对齐模型行为恢复安全奖励并在解码时反转
Published Oct 6
Evaluated models Llama3.1-8B-Instruct , Gemma-7B-Instruct , Qwen2.5-7B-Instruct and others9 Abstract ReSA 用最大熵 IRL 从对齐模型行为恢复代理安全奖励，并在解码时反转以诱导有害输出。ReSA 是一种对抗攻击：从对齐模型的可观察行为恢复代理安全奖励，再在解码时把该奖励反过来用。 。
提出只改 SAE 解码器的后门，在冻结语言模型上植入代码插入
Published Oct 5
Evaluated models Gemma 3 4B Instruct , Qwen3 1.7B , Llama 3.1 8B Instruct Abstract 只改解码器的 SAE 可带上代码插入后门，且若干常规质量指标变化不大。只改 SAE 解码器、冻结编码器与语言模型，即可把代码插入行为放进推理时的辅助组件。作者把这视为 SAE 供应链问题：checkpoint 一旦替换某一层激活，就参与后续计算，而不只是解释工具。
提出生成时防御 SENTINEL，匹配输入输出以抽出越狱意图并停写
Published Oct 3
Evaluated models Llama2-7B , Llama3-8B , Mistral-7B-v0.2 and others6 用分类后缀提升激活探针对分布外恶意输入的检测
Published Oct 2
Evaluated models Llama-3.1-8B , Qwen3.5-9B , Gemma-4-12B Abstract 作者称排序增益来自分类格式，低 FPR 精度来自点名准则。用户轮后的分类后缀，被作者用作激活探针的分布外检测干预：排序看格式，低 FPR 精度看是否点名恶意/良性。
提出 NDDL，用良性扩散轨迹检测文生图后门提示并定位触发 token
Published Sep 30
Evaluated models Stable Diffusion v1.5 , Stable Diffusion XL , Stable Diffusion v3.5 and others4 Abstract NDDL 从良性扩散轨迹学习正常转移，用动态不一致检测后门并定位触发 token。NDDL 是面向文生图扩散模型的后门防御：防御者只有白盒模型访问和良性提示词，没有后门先验，需要区分后门提示词并找出触发 token。
提出 NEEDLE，用权重正交化在已知触发器时去除大模型后门
Published Sep 30
Evaluated models Gemma-3-4B-IT , Qwen3-4B-Instruct-2507 , Gemma-3-12B-IT Abstract NEEDLE 用闭式权重编辑去掉已知触发器的后门，并保住拒绝子空间。
提出 Constitutional adapters，蒸馏宪法行为以缓解越狱与失准
Published Sep 29
Evaluated models Qwen3.5-4B , Llama-3.1-8B-Instruct , Gemma-4-E4B and others4 Abstract 差分宪法适配器在长上下文和多轮上更稳，并可按 α 换取遵从。
诊断护栏相对基座在对抗提示上的置信失校
Published Sep 29
Evaluated models Llama-Guard 7B , Llama-Guard-2 8B , Llama-Guard-3 8B and others10 Abstract 对抗包装使 guard 高置信漏检，而对应 base 在同一输入上常更不确定。诊断性分析 ：五个提示词分类 guard 的置信度，在清洁输入上可以较校准，在越狱包装后则变成高置信度的错误判定。
提出 CAM-Steer，推理时按多类别风险引导隐藏状态以降低有害回复
Published Sep 28
Evaluated models Qwen3-8B , Llama-3.1-8B-Instruct , Gemma-2-9B-IT 提出 AnchorRep，用 LoRA 推离有害提示表征以防御跨模型越狱迁移
Published Sep 26
Evaluated models Llama-3-8B , Mistral-7B , Vicuna-7B and others5 用能量势垒信号检测扩散语言模型的越狱
Published Sep 25
Evaluated models LLaDA-8B-Instruct , LLaDA-1.5 , Dream-v0-Instruct-7B and others4 Abstract 能量势垒把 dLLM 越狱分成掩盖初始状态或中途越垒，三个 logit 信号据此互补检测。Bach 等提出用去噪能量景观解释 masked diffusion 语言模型上的越狱，并由此得到三个免训练检测信号。 。
提出 AEGIS，用中层内部信号选择性激活后层适配器拦截音频越狱
Published Sep 24
Evaluated models Qwen2-Audio-7B , VITA-1.5 , Ultravox v0.5 and others6 提出 PALS，用潜空间平滑防御全双工语音对话的对抗扰动
Published Sep 23
Evaluated models Moshi-style agent