Folgen
Mingze Wang
Mingze Wang
School of Mathematical Sciences, Peking University
Bestätigte E-Mail-Adresse bei stu.pku.edu.cn - Startseite
Titel
Zitiert von
Zitiert von
Jahr
The Alignment Property of SGD Noise and How it Helps Select Flat Minima: A Stability Analysis
L Wu, M Wang, WJ Su
Advances in Neural Information Processing Systems (NeurIPS 2022), 1-25, 2022
33*2022
Generalization Error Bounds for Deep Neural Networks Trained by SGD
M Wang, C Ma
arXiv: 2206.03299, 1-32, 2022
72022
Understanding Multi-phase Optimization Dynamics and Rich Nonlinear Behaviors of ReLU Networks
M Wang, C Ma
Advances in Neural Information Processing Systems (NeurIPS 2023, Spotlight …, 2023
62023
A Theoretical Analysis of Noise Geometry in Stochastic Gradient Descent
M Wang, L Wu
NeurIPS 2023 Workshop on M3L, 1-30, 2023
3*2023
Early Stage Convergence and Global Convergence of Training Mildly Parameterized Neural Networks
M Wang, C Ma
Advances in Neural Information Processing Systems (NeurIPS 2022), 1-73, 2022
32022
The Implicit Bias of Gradient Noise: A Symmetry Perspective
L Ziyin, M Wang, L Wu
arXiv: 2402.07193, 1-17, 2024
2024
Understanding the Expressive Power and Mechanisms of Transformer for Sequence Modeling
M Wang, W E
arXiv: 2402.00522, 1-65, 2024
2024
Achieving Margin Maximization Exponentially Fast via Progressive Norm Rescaling
M Wang, Z Min, L Wu
International Conference on Machine Learning (ICML 2024), 1-38, 2023
2023
Das System kann den Vorgang jetzt nicht ausführen. Versuchen Sie es später erneut.
Artikel 1–8