Explainable and Fine-Grained Safeguarding of LLM Multi-Agent Systems via Bi-Level Graph Anomaly Detection
通过双层图异常检测实现LLM多智能体系统的可解释性与细粒度防护
机构 * School of Information and Communication Technology, Griffith University, Australia(格里菲斯大学信息与通信技术学院) ; School of Artificial Intelligence, Jilin University, China(吉林大学人工智能学院) ; Department of Statistics and Data Science, Northwestern University, USA(西北大学统计与数据科学系)
专题命中 知识编辑与模型理解 :LLM(title,abstract);large language model(abstract);language model(abstract);分类 cs.AI
AI总结 本文提出XG-Guard框架,通过双层图异常检测实现对LLM多智能体系统中恶意智能体的可解释性与细粒度检测。
Comments 14 pages, 3 tables, 5 figures