2 papers
cs.LG2023
A Theoretical Explanation of Activation Sparsity through Flat Minima and Adversarial Robustness
Ze Peng, Lei Qi, Yinghuan Shi +1
A recent empirical observation (Li et al., 2022b) of activation sparsity in MLP blocks offers an opportunity to drastically reduce computation costs for free. Although having attri…
cs.CV2023
Enhancing Sample Utilization through Sample Adaptive Augmentation in Semi-Supervised Learning
Guan Gui, Zhen Zhao, Lei Qi +3
In semi-supervised learning, unlabeled samples can be utilized through augmentation and consistency regularization. However, we observed certain samples, even undergoing strong aug…