Backdoor Attacks on Pre-trained Models by Layerwise Weight PoisoningDownload PDFOpen Website

Linyang Li, Demin Song, Xiaonan Li, Jiehang Zeng, Ruotian Ma, Xipeng Qiu

2021 (modified: 20 Nov 2021)EMNLP (1) 2021Readers: Everyone
Abstract: Linyang Li, Demin Song, Xiaonan Li, Jiehang Zeng, Ruotian Ma, Xipeng Qiu. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. 2021.
0 Replies