Philipp Zimmermann
← Back to Newsletter

Paper Library

Collection of AI Security research papers

Showing 1331 papers total

September 15 - September 21, 2025

16 papers

SABER: Uncovering Vulnerabilities in Safety Alignment via Cross-Layer Residual Connection

Maithili Joshi, Palash Nandi, Tanmoy Chakraborty
2025-09-19
red teaming
2509.16060v1

EmoQ: Speech Emotion Recognition via Speech-Aware Q-Former and Large Language Model

Yiqing Yang, Man-Wai Mak
2025-09-19
2509.15775v1

Beyond Surface Alignment: Rebuilding LLMs Safety Mechanism via Probabilistically Ablating Refusal Direction

Yuanbo Xie, Yingjie Zhang, Tianyun Liu, Duohe Ma, Tingwen Liu
2025-09-18
red teaming safety
2509.15202v1

Sentinel Agents for Secure and Trustworthy Agentic AI in Multi-Agent Systems

Diego Gosmar, Deborah A. Dahl
2025-09-18
red teaming
2509.14956v1

Toxicity Red-Teaming: Benchmarking LLM Safety in Singapore's Low-Resource Languages

Yujia Hu, Ming Shan Hee, Preslav Nakov, Roy Ka-Wei Lee
2025-09-18
safety
2509.15260v2

Toxicity Red-Teaming: Benchmarking LLM Safety in Singapore's Low-Resource Languages

Yujia Hu, Ming Shan Hee, Preslav Nakov, Roy Ka-Wei Lee
2025-09-18
safety
2509.15260v1

MUSE: MCTS-Driven Red Teaming Framework for Enhanced Multi-Turn Dialogue Safety in Large Language Models

Siyu Yan, Long Zeng, Xuecheng Wu, Chengcheng Han, Kongcheng Zhang, Chong Peng, Xuezhi Cao, Xunliang Cai, Chenjuan Guo
2025-09-18
red teaming
2509.14651v1

LLM Jailbreak Detection for (Almost) Free!

Guorui Chen, Yifan Xia, Xiaojun Jia, Zhijiang Li, Philip Torr, Jindong Gu
2025-09-18
red teaming
2509.14558v1

A Simple and Efficient Jailbreak Method Exploiting LLMs' Helpfulness

Xuan Luo, Yue Wang, Zefeng He, Geng Tu, Jing Li, Ruifeng Xu
2025-09-17
red teaming
2509.14297v1

Agentic JWT: A Secure Delegation Protocol for Autonomous AI Agents

Abhishek Goswami
2025-09-16
2509.13597v1

A Multi-Agent LLM Defense Pipeline Against Prompt Injection Attacks

S M Asif Hossain, Ruksat Khan Shayoni, Mohd Ruhul Ameen, Akif Islam, M. F. Mridha, Jungpil Shin
2025-09-16
2509.14285v2

A Multi-Agent LLM Defense Pipeline Against Prompt Injection Attacks

S M Asif Hossain, Ruksat Khan Shayoni, Mohd Ruhul Ameen, Akif Islam, M. F. Mridha, Jungpil Shin
2025-09-16
safety
2509.14285v1

Jailbreaking Large Language Models Through Content Concretization

Johan Wahréus, Ahmed Hussain, Panos Papadimitratos
2025-09-16
red teaming
2509.12937v1

Defense-to-Attack: Bypassing Weak Defenses Enables Stronger Jailbreaks in Vision-Language Models

Yunhan Zhao, Xiang Zheng, Xingjun Ma
2025-09-16
red teaming
2509.12724v1

Early Approaches to Adversarial Fine-Tuning for Prompt Injection Defense: A 2022 Study of GPT-3 and Contemporary Models

Gustavo Sandoval, Denys Fenchenko, Junyao Chen
2025-09-15
red teaming
2509.14271v1

Reasoned Safety Alignment: Ensuring Jailbreak Defense via Answer-Then-Check

Chentao Cao, Xiaojun Xu, Bo Han, Hang Li
2025-09-15
2509.11629v1

September 08 - September 14, 2025

8 papers

Securing AI Agents: Implementing Role-Based Access Control for Industrial Applications

Aadil Gani Ganie
2025-09-14
2509.11431v1

When Smiley Turns Hostile: Interpreting How Emojis Trigger LLMs' Toxicity

Shiyao Cui, Xijia Feng, Yingkang Wang, Junxiao Yang, Zhexin Zhang, Biplab Sikdar, Hongning Wang, Han Qiu, Minlie Huang
2025-09-14
red teaming
2509.11141v1

ENJ: Optimizing Noise with Genetic Algorithms to Jailbreak LSMs

Yibo Zhang, Liang Lin
2025-09-14
2509.11128v1

Harmful Prompt Laundering: Jailbreaking LLMs with Abductive Styles and Symbolic Encoding

Seongho Joo, Hyukhun Koh, Kyomin Jung
2025-09-13
red teaming
2509.10931v1

Prompt Injection Attacks on LLM Generated Reviews of Scientific Publications

Janis Keuper
2025-09-12
red teaming
2509.10248v3

Realism Control One-step Diffusion for Real-World Image Super-Resolution

Zongliang Wu, Siming Zheng, Peng-Tao Jiang, Xin Yuan
2025-09-12
2509.10122v2

When Your Reviewer is an LLM: Biases, Divergence, and Prompt Injection Risks in Peer Review

Changjia Zhu, Junjie Xiong, Renkai Ma, Zhicong Lu, Yao Liu, Lingyao Li
2025-09-12
red teaming
2509.09912v1

Steering MoE LLMs via Expert (De)Activation

Mohsen Fayyaz, Ali Modarressi, Hanieh Deilamsalehy, Franck Dernoncourt, Ryan Rossi, Trung Bui, Hinrich Schütze, Nanyun Peng
2025-09-11
red teaming
2509.09660v1