Stateful Cooperative Agents Safeguarding LLMs Against Evolving Multi-Turn Attacks
Siyuan Li, Zehao Liu, Haoyu Li, Xi Lin, Ning Liu, Jun Wu, Jianhua Li, Mohsen Guizani
cs.CR
Submitted: 2026-07-31
Code: https://github.com/SiyuanLi00/CoopGuard
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- GPT-4 Technical Report
- Gemini: A Family of Highly Capable Multimodal Models
- LLaMA: Open and Efficient Foundation Language Models
- A Survey of Large Language Models in Medicine: Progress, Application, and Challenge
- Exploring Recommendation Capabilities of GPT-4V(ision): A Preliminary Case Study
- Exploiting the Index Gradients for Optimization-Based Jailbreaking on Large Language Models
- Helping Large Language Models Protect Themselves: An Enhanced Filtering and Summarization System
- X-Teaming: Multi-Turn Jailbreaks and Defenses with Adaptive Multi-Agents
- Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation
- DataSentinel: A Game-Theoretic Detection of Prompt Injection Attacks
- MasterKey: Automated Jailbreak Across Multiple Large Language Model Chatbots
- Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow Instructions
- Attacking Large Language Models with Projected Gradient Descent
- Universal and Transferable Adversarial Attacks on Aligned Language Models
- AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models
- AdvPrompter: Fast Adaptive Adversarial Prompting for LLMs
- How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMs
- GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts
- HoneyTrap: Deceiving Large Language Model Attackers to Honeypot Traps with Resilient Multi-Agent Defense
- Great, Now Write an Article About That: The Crescendo Multi-Turn LLM Jailbreak Attack
Related papers
- SoK: AI-Augmented Binary Reversing
- Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
- Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
- Efficient Fuzzy PSI under One-Sided Assumptions
- Sealing the Audit-Runtime Gap for LLM Skills
- Token Composition: A Graph Based on EVM Logs