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

Oct 13 · Tuesday

All papers

Found 5 papers

Filter5

Attack type

Jailbreaks38 Prompt injection4 Poisoning and backdoors4 Detection evasion5 Malicious use0

Attack technique

Gradient and representation optimization9 Automated search5 Encoding and obfuscation5 Cross-modal payloads3 Defense-targeted14

System type

Models and APIs5 coding agent0

Attack surface / entry point

User input5

Layer

Prompt layer4

System component

Guardrails and judges3

Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 227papers have no tags yet and are not included in filtered results.

Attack type：Detection evasion Attack type：Malicious use Attack technique：Encoding and obfuscation System type：Models and APIs System type：coding agent Clear all

Another 227papers do not have category tags yet and are not included in filtered results.

InterpretabilityarXiv 2609.24801Sep 22
研究用 XAI 归因分析 Prompt Guard 2，同义词替换可翻转其提示注入判定
用词级归因指导同义词替换，翻转提示注入检测并验证越狱
arXiv
2609.24801
Published
Sep 22
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
Prompt Guard 2 (86M), Llama 3.1 8B
- DefensesDetection and filtering
- AttacksDetection evasion
- TechniqueGradient and representation optimization
- ComponentGuardrails and judges
Abstract作者称检测靠分散的词法线索。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2609.05843Sep 5
SinoGlyphBench：中文字形混淆下语言模型审核的诊断基准
提出 SinoGlyphBench，用配对字形混淆诊断中文审核稳定性
arXiv
2609.05843
Published
Sep 5
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
GPT-5.4 mini, GPT-5.5, DeepSeek V4 Flash and others12
- AttacksDetection evasion
- TechniqueEncoding and obfuscation
- ComponentGuardrails and judges
AbstractSinoGlyphBench 显示中文字形混淆会同时增加有害漏报和误报，全文扰动最重。
SinoGlyphBench是一个配对诊断基准，用来看内容审核在可还原的中文字形混淆下是否还保持原决定。
Paper analysisRead analysis
AttacksarXiv 2605.07481May 8
Vaporizer：研究者用改写类攻击破解大语言模型输出水印方案
提出 Vaporizer，用词法变换、翻译和释义破解 LLM 输出水印
arXiv
2605.07481
Published
May 8
Layer
Prompt layer
Setting
Models and APIs
- AttacksDetection evasion
- TechniqueEncoding and obfuscation
Paper analysisRead analysis
AttacksarXiv 2603.08104Mar 9
Invisible Security Threat: Malicious Fine-Tuning of LLMs via Steganography
提出隐写恶意微调，使模型界面显良性并隐蔽传出有害内容
arXiv
2603.08104
Published
Mar 9
Layer
Training and data layer
Setting
Models and APIs
Attacker
White-box
Evaluated models
GPT-4.1-2025-04-14, Llama-3.3-70B-Instruct, Phi-4 and others7
- AttacksModel tampering
- TechniqueEncoding and obfuscation
- ComponentFine-tuning and open weights
Paper analysisRead analysis
AttacksarXiv 2509.09112Sep 11, 2025
研究：字符级扰动可破坏 LLM 水印，并提出基于遗传算法的去除攻击
提出字符级扰动与遗传算法去除 LLM 水印
arXiv
2509.09112
Published
Sep 11, 2025
Layer
Prompt layer
Setting
Models and APIs
Attacker
Black-box
Evaluated models
OPT-1.3B, LLaMA-3-8B
- AttacksDetection evasion
- TechniqueEncoding and obfuscation
Abstract字符级扰动靠更大攻击范围移除水印。
这篇工作评估推理时 LLM 水印在黑盒、有限查询下的移除难度，并比较字符级、token 级和句子级编辑。
Paper analysisRead analysis

You have reached the end

AI Safety Daily

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

All papers

Attack type

Attack technique

System type

Attack surface / entry point

Layer

System component

研究用 XAI 归因分析 Prompt Guard 2，同义词替换可翻转其提示注入判定

SinoGlyphBench：中文字形混淆下语言模型审核的诊断基准

Vaporizer：研究者用改写类攻击破解大语言模型输出水印方案

Invisible Security Threat: Malicious Fine-Tuning of LLMs via Steganography

研究：字符级扰动可破坏 LLM 水印，并提出基于遗传算法的去除攻击