Shen Dong — accepted AI-conference papers

2025

Memory Injection Attacks on LLM Agents via Query-Only Interaction

Shen Dong, Shaochen Xu, Pengfei He, Yige Li, Jiliang Tang, Tianming Liu +2

NeurIPS 2025poster

Agents powered by large language models (LLMs) have demonstrated strong capabilities in a wide range of complex, real-world applications. However, LLM agents with a compromised memory bank may easily produce harmful outputs when the past records retrieved for demonstration are malicious. In this pap…

Cited by 0 Source ↗Scholar ↗

2025

Red-Teaming LLM Multi-Agent Systems via Communication Attacks

Pengfei He, Yuping Lin, Shen Dong, Han Xu, Yue Xing, Hui Liu

ACL 2025finding

Large Language Model-based Multi-Agent Systems (LLM-MAS) have revolutionized complex problem-solving capability by enabling sophisticated agent collaboration through message-based communications. While the communication framework is crucial for agent coordination, it also introduces a critical yet u…

Cited by 0 Source ↗PDF ↗Scholar ↗