Knowledge distillation is normally used to compress a big network, or teacher, onto a smaller one, the student, by training it to match its outputs. Recently, some works have …
While adversarial training and its variants have shown to be the most effective algorithms to defend against adversarial attacks, their extremely slow training process makes it hard to …
Ensuring the trustworthiness of large language models (LLMs) is crucial. Most studies concentrate on fully pre-trained LLMs to better understand and improve LLMs' …
J Chen, Y Cao, Q Gu - Uncertainty in Artificial Intelligence, 2023 - proceedings.mlr.press
Benign overfitting, where classifiers memorize noisy training data yet still achieve a good generalization performance, has drawn great attention in the machine learning community …
Y Liu, C Wang, X Yuan - arXiv preprint arXiv:2406.12222, 2024 - arxiv.org
Federated Learning (FL) is susceptible to poisoning attacks, wherein compromised clients manipulate the global model by modifying local datasets or sending manipulated model …
X Cheng, H Zhang, Y Xin, W Shen, J Ren… - arXiv preprint arXiv …, 2022 - arxiv.org
This paper mathematically derives an analytic solution of the adversarial perturbation on a ReLU network, and theoretically explains the difficulty of adversarial training. Specifically …
Deep learning has been able to outperform humans in terms of classification accuracy in many tasks. However, to achieve robustness to adversarial perturbations, the best …