Abstract: In recent years, attention-based models have excelled across various domains but remain vulnerable to backdoor attacks, often from downloading or fine-tuning on poisoned datasets. Many ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback