Haobin Ke
1 indexed paper
Recent (6 mo)
1With code
0Influential cites
0Benchmarked
0Publications per year
126
Top categories
Crypto×1NLP×1ML×1
Frequent co-authors
Research Timeline
2026
When Routine Chats Turn Toxic: Unintended Long-Term State Poisoning in Personalized Agents
The paper identifies 'unintended long-term state poisoning'—a security risk where routine user interactions gradually corrupt an LLM agent's persistent state—and proposes a defense mechanism called StateGuard.
Highlighted terms show continued research focus across papers
Papers
cs.CRcs.CLcs.LGRecentMay 7, 2026
When Routine Chats Turn Toxic: Unintended Long-Term State Poisoning in Personalized Agents
Xiaoyu Xu, Minxin Du, Qipeng Xie, Haobin Ke +2 more
The paper identifies 'unintended long-term state poisoning'—a security risk where routine user interactions gradually corrupt an LLM agent's persistent state—and proposes a defense mechanism called St…
View →