Translation in progress Some entries still include text in the original language.
Oct 12 · Monday
All papers Found 28 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.
Uses premise permutation and hidden-layer tracing to audit whether legal chain-of-thought truly relies on the cited legal provisions.
Published Oct 9
Evaluated models Qwen3-8B , Qwen3-14B , Llama-3.1-8B and others7 Abstract Using counterfactual substitution of authorities and internal commitment tracing, the authors find that legal CoT rulings depend on authorities only to a limited extent and are vulnerable to prompt injection manipulation.This study conducts an empirical audit of the causal truthfulness of legal citations in chain-of-thought reasoning of large language models in the legal domain.
Proposes the semantic watermark SemField, which still identifies the attribution of generated text after paraphrasing and structural changes.
Published Oct 8
Evaluated models Llama-3.2-1B , Qwen3-1.7B , OPT-1.3B Abstract Training-independent document-level semantic watermark.SemField is a zero-bit semantic watermark embedded by the provider at generation time and detected later with a key, used to identify the provider's own model text after paraphrasing and sentence-level structural changes.
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。
构建多语言音视频 DeepFake 基准 BabelFake 并评测检测器
Published Oct 5
Evaluated models SpeechForensics , AVH-Align , (B)RAVEn Abstract BabelFake 为五语同意采集基准。BabelFake 是经同意与 IRB 批准采集的多语音视频 DeepFake 检测基准,用来把视觉操纵和语音合成拆开比较。
把 SynthID 嵌入基因组语言模型采样,并用位置无关检测追溯生成序列
Published Oct 4
Evaluated models Carbon-500M , GENERator-v2-eukaryote-1.2b-base Abstract 两个六碱基基因组模型嵌入 SynthID 后,位置无关验证能找回水印,作者称这只是统计检测。作者在两个六碱基基因组模型的采样里嵌入 SynthID,并让验证者在不知道起点、链和 token phase 时做一次校正后的检出。
提出 THRIVE,为自回归 TTS 嵌入抗重建的多比特语音水印
Published Oct 3
Evaluated models Qwen3-TTS (0.6B) , Dots.tts (2B) 提出 LiBRA,在潜空间双向优化以规避图像水印双侧检测
Published Oct 2
Evaluated models Stable Signature , YU1 , HiDDeN and others4 Abstract LiBRA 用双向潜空间优化规避双侧水印检测,并提高相对反转基线的保真。
提出抗删除水印 RMCW,在生成时注入 Reed-Muller 结构并做局部检验
Published Oct 2
Evaluated models OPT-1.3B , Llama-3.1-8B-Instruct 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 UNTAG,绕过开源权重模型的 trigger-tag 滥用检测
Published Oct 2
Evaluated models Llama2 (7B) , Llama3 (8B) , Qwen2.5 (7B) Abstract 作者用 UNTAG 评估 trigger-tag,认为其不能充当开放权重滥用检测的主要边界。UNTAG 是针对开放权重 LLM 上条件滥用检测 trigger-tag 的统一攻击评估:信号可以在解码期注入,也可以写进权重,但攻击者若能改输出或改已发布权重,检测可以失效。
提出 TAILOR,按请求组合图像水印以满足抗攻击与画质延迟规格
Published Oct 1 用因果激活修补定位模型服从提示注入的决策层
Published Sep 29
Evaluated models Qwen3-4B , Qwen3-14B , Qwen3-32B and others7 Abstract 合规的因果杠杆集中在网络后三分之一的低秩子空间,该层也是混淆下检测最好的位置。这篇工作用逐层激活修补定位提示注入合规发生在网络何处,并把该位置用于检测验证。 。
提出 AutoMark,自动搜索并统计检验 LLM 文本水印
Published Sep 29
Evaluated models Llama-3.1-8B-Instruct , Qwen3.8-27B , GPT-6 Astra and others5 Abstract AutoMark 的贡献同时包含搜索出的水印,以及用于约束自动研究误报与评测漏洞的验证框架。AutoMark 是一套自动水印研究框架 :先检查生成分布和误报性质,再比较检测、编辑鲁棒性与质量,最后人工分析组件并做留出评测。
比较用于审计提示注入与隐瞒的 token 位置选择信号
Published Sep 29
Evaluated models Qwen2.5-7B , Gemma-3-12B , Gemma-3-27B and others4 提出成对无失真水印 TTMARK,增强低熵下的机器文本检测
Published Sep 29
Evaluated models Llama3.2-3B-Instruct , Mistral-v0.3-7B-Instruct , Qwen2.5-7B-Instruct Abstract TTMARK 用水印化相邻 token 对利用条件熵,在多种无失真方案上提高检测并保持生成质量。TTMARK 是一种成对、无失真的文本水印:对相邻 token 的联合分布重加权,而不是逐 token 独立重加权。
提出带密钥的 SNW,抵御事后图像水印的身份伪造
Published Sep 28
Evaluated models SNW Abstract SNW 用密钥与满秩各向同性投影提高事后水印的 Shannon 容量,并对 PCA 完全安全。Gesny 与 Giboulot 针对生成内容追溯所依赖的多比特事后图像水印,批评端到端 DNN 方案既无容量理论、又无密钥。
提出 VidMark 水印方案,核验视频生成模型的输出来源与所有权
Published Sep 20
Evaluated models Stable Video Diffusion (SVD) , Open-Sora , Wan Abstract VidMark 加解码器引导微调,把模型签名写入 VGM 输出并用于两项核验。Huang 等人提出一套面向视频生成模型的生成内嵌水印方案,用来同时判断一条视频是否来自受保护模型,以及一个嫌疑模型是否为该模型的拷贝。
提出 DenMark,在扩散语言模型去噪时注入可抵抗保义编辑的语义水印
Published Sep 13
Evaluated models LLaDA-8B , LLaDA1.5-8B , LLaDA2.0-mini and others4 Abstract DenMark 用 rollout 在 DLM 去噪中累积语义水印,扫描检测以抵抗保义编辑。DenMark 是面向扩散语言模型的语义水印:在去噪的局部更新里注入密钥相关语义信号,并在编辑改变 token 边界后仍检测。现有句子级语义水印要先完成整个语义单元再打分取舍,而 DLM 以任意顺序更新掩码,中间单元不完整。DenMark 把续写划成固定 token 区域,对每个局部候选做不提交的 rollout 补全,按语义分 gb 的平均选择更新,从而沿轨迹累积证据。
复现 Tree-Ring 语义水印的黑盒伪造并恢复检测统计量
Published Sep 11
Evaluated models Stable Diffusion XL base 1.0 提出黑盒攻击 DRIFT,偏转生成轨迹以去除扩散水印
Published Sep 8