AutoDefense: Multi-Agent LLM Defense against Jailbreak Attacks

被引：0
|
作者：
Zeng, Yifan [1 ]
Wu, Yiran [2 ]
Zhang, Xiao [3 ]
Wang, Huazheng [1 ]
Wu, Qingyun [2 ]
机构：
[1] Oregon State University, United States
[2] Pennsylvania State University, United States
[3] CISPA Helmholtz Center for Information Security, Germany
来源：
arXiv |
关键词：
Agent systems - Filtering mechanism - Language model - Large models - Model agents - Multi agent - Open-source - Performance - Pre-training;
D O I：
暂无
中图分类号：
学科分类号：
摘要：
56
引用
收藏
相关论文
共 50 条
  • [1] The software environment for multi-agent simulation of defense mechanisms against DDoS attacks
    Kotenko, Igor
    Ulanov, Alexander
    INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE FOR MODELLING, CONTROL & AUTOMATION JOINTLY WITH INTERNATIONAL CONFERENCE ON INTELLIGENT AGENTS, WEB TECHNOLOGIES & INTERNET COMMERCE, VOL 1, PROCEEDINGS, 2006, : 283 - +
  • [2] Multi-agent framework for simulation of adaptive cooperative defense against Internet attacks
    Kotenko, Igor
    Ulanov, Alexander
    AUTONOMOUS INTELLIGENT SYSTEMS: AGENTS AND DATA MINING, PROCEEDINGS, 2007, 4476 : 212 - +
  • [3] A Novel Defense Strategy Against Zero-Dynamics Attacks in Multi-Agent Systems
    Mao, Yanbing
    Akyol, Emrah
    Zhang, Ziang
    2019 IEEE 58TH CONFERENCE ON DECISION AND CONTROL (CDC), 2019, : 3563 - 3568
  • [4] Defense methods against multi-language and multi-intent LLM attacks
    Fan, Sunjia
    Yang, Yichao
    Huang, Weiqi
    Ma, Ke
    Liu, Yucen
    Zheng, Yuxin
    Proceedings of SPIE - The International Society for Optical Engineering, 2024, 13403
  • [5] AutoDefense: Reinforcement Learning Based Autoreactive Defense Against Network Attacks
    Mi, Yu
    Mohaisen, David
    Wang, An
    2022 IEEE CONFERENCE ON COMMUNICATIONS AND NETWORK SECURITY (CNS), 2022, : 163 - 171
  • [6] A defense strategy for false data injection attacks in multi-agent systems
    Sun, Lucheng
    Wu, Tiejun
    Zhang, Ya
    INTERNATIONAL JOURNAL OF SYSTEMS SCIENCE, 2023, 54 (16) : 3071 - 3084
  • [7] Break the Breakout: Reinventing LM Defense Against Jailbreak Attacks with Self-Refinement
    Kim, Heegyu
    Yuk, Sehyun
    Cho, Hyunsouk
    arXiv, 1600,
  • [8] A Multi-Agent Intelligence Hybrid System Technique for Detection and Defense of DDoS Attacks
    Chen, Hsia-Hsiang
    Huang, Shih-Kun
    INTELLIGENT SYSTEMS AND APPLICATIONS (ICS 2014), 2015, 274 : 125 - 139
  • [9] MARNet: Backdoor Attacks Against Cooperative Multi-Agent Reinforcement Learning
    Chen, Yanjiao
    Zheng, Zhicong
    Gong, Xueluan
    IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING, 2023, 20 (05) : 4188 - 4198
  • [10] Security Analysis of Poisoning Attacks Against Multi-agent Reinforcement Learning
    Xie, Zhiqiang
    Xiang, Yingxiao
    Li, Yike
    Zhao, Shuang
    Tong, Endong
    Niu, Wenjia
    Liu, Jiqiang
    Wang, Jian
    ALGORITHMS AND ARCHITECTURES FOR PARALLEL PROCESSING, ICA3PP 2021, PT I, 2022, 13155 : 660 - 675
12345