Cheng Liu, Xiaolei Liu, Xingyu Li, Bangzhou Xin, Kangyi Ding. TrajGuard: Streaming Hidden-state Trajectory Detection for Decoding-time Jailbreak Defense. In Maria Liakata, Viviane P. Moreira, Jiajun Zhang 0001, David Jurgens, editors, Findings of the Association for Computational Linguistics, ACL 2026, San Diego, California, United States, July 2-7, 2026. pages 13371-13388, Association for Computational Linguistics, 2026. [doi]
Abstract is missing.