大语言模型微调阶段的后门投毒：攻击、防御与评估
马文飞, 龙春, 李婧, 杨帆, 付豫豪
Backdoor poisoning in the fine-tuning stage of large language models: attacks, defenses, and evaluation
Ma Wenfei, Long Chun, Li Jing, Yang Fan, Fu Yuhao
信息网络安全
.
2026, (9): 1354
-1368
.
DOI: 10.3969/j.issn.1671-1122.2026.09.003