May I have your Attention? Breaking Fine-Tuning based Prompt Injection Defenses using Architecture-Aware Attacks
Authors: Nishit V. Pandya, Andrey Labunets, Sicun Gao, Earlence Fernandes | Published: 2025-07-10
敵対的攻撃
BarkBeetle: Stealing Decision Tree Models with Fault Injection
Authors: Qifan Wang, Jonas Sander, Minmin Jiang, Thomas Eisenbarth, David Oswald | Published: 2025-07-09
CAVGAN: Unifying Jailbreak and Defense of LLMs via Generative Adversarial Attacks on their Internal Representations
Authors: Xiaohu Li, Yunfeng Ning, Zepeng Bao, Mayi Xu, Jianhao Chen, Tieyun Qian | Published: 2025-07-08
The Hidden Threat in Plain Text: Attacking RAG Data Loaders
Authors: Alberto Castagnaro, Umberto Salviati, Mauro Conti, Luca Pajola, Simeone Pizzi | Published: 2025-07-07
Amplifying Machine Learning Attacks Through Strategic Compositions
Authors: Yugeng Liu, Zheng Li, Hai Huang, Michael Backes, Yang Zhang | Published: 2025-06-23
LLMs Cannot Reliably Judge (Yet?): A Comprehensive Assessment on the Robustness of LLM-as-a-Judge
Authors: Songze Li, Chuokun Xu, Jiaying Wang, Xueluan Gong, Chen Chen, Jirui Zhang, Jun Wang, Kwok-Yan Lam, Shouling Ji | Published: 2025-06-11
A Cryptographic Perspective on Mitigation vs. Detection in Machine Learning
Authors: Greg Gluch, Shafi Goldwasser | Published: 2025-04-28 | Updated: 2025-07-10
Support is All You Need for Certified VAE Training
Authors: Changming Xu, Debangshu Banerjee, Deepak Vasisht, Gagandeep Singh | Published: 2025-04-16
Language Models May Verbatim Complete Text They Were Not Explicitly Trained On
Authors: Ken Ziyu Liu, Christopher A. Choquette-Choo, Matthew Jagielski, Peter Kairouz, Sanmi Koyejo, Percy Liang, Nicolas Papernot | Published: 2025-03-21 | Updated: 2025-03-25
Immune: Improving Safety Against Jailbreaks in Multi-modal LLMs via Inference-Time Alignment
Authors: Soumya Suvra Ghosal, Souradip Chakraborty, Vaibhav Singh, Tianrui Guan, Mengdi Wang, Ahmad Beirami, Furong Huang, Alvaro Velasquez, Dinesh Manocha, Amrit Singh Bedi | Published: 2024-11-27 | Updated: 2025-03-20