Wenqian Yu 0001, Shuo Chen, Zhijiang Li, Zhipeng Wang, Jindong Gu. You Only Need One Single Token to Refine Safety Alignment. 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 13529-13545, Association for Computational Linguistics, 2026. [doi]
Abstract is missing.