中文版 | English

在结果中检索

CASIA OpenIR

浏览/检索结果: 共2条，第1-2条 帮助

已选(0)清除 条数/页：   排序方式：
Red Alarm for Pre-trained Models: Universal Vulnerability to Neuron-level Backdoor Attacks 期刊论文
Machine Intelligence Research, 2023, 卷号: 20, 期号: 2, 页码: 180-193
作者:  Zhengyan Zhang;  Guangxuan Xiao;  Yongwei Li;  Tian Lv;  Fanchao Qi;  Zhiyuan Liu;  Yasheng Wang;  Xin Jiang;  Maosong Sun
Adobe PDF(1649Kb)  |  收藏  |  浏览/下载：16/3  |  提交时间：2024/04/23
Pre-trained language models  backdoor attacks  transformers  natural language processing (NLP)  computer vision (CV)  
Zero-shot language extension for dialogue state tracking via pre-trained models and multi-auxiliary-tasks fine-tuning 期刊论文
KNOWLEDGE-BASED SYSTEMS, 2023, 卷号: 259, 页码: 14
作者:  Xiang, Lu;  Zhao, Yang;  Zhu, Junnan;  Zhou, Yu;  Zong, Chengqing
收藏  |  浏览/下载：139/0  |  提交时间：2023/03/20
Dialogue state tracking  Zero -shot language extension  Multilingual DST  Pre -trained models  Multi -auxiliary -tasks fine-tuning