Следене
Wei Hu
Wei Hu
Assistant Professor of Computer Science and Engineering, University of Michigan
Потвърден имейл адрес: umich.edu - Начална страница
Заглавие
Позовавания
Позовавания
Година
Fine-grained analysis of optimization and generalization for overparameterized two-layer neural networks
S Arora, S Du, W Hu, Z Li, R Wang
International conference on machine learning, 322-332, 2019
11482019
On exact computation with an infinitely wide neural net
S Arora, SS Du, W Hu, Z Li, RR Salakhutdinov, R Wang
Advances in neural information processing systems 32, 2019
10692019
Implicit regularization in deep matrix factorization
S Arora, N Cohen, W Hu, Y Luo
Advances in neural information processing systems 32, 2019
6202019
A convergence analysis of gradient descent for deep linear neural networks
S Arora, N Cohen, N Golowich, W Hu
arXiv preprint arXiv:1810.02281, 2018
3062018
Few-shot learning via learning the representation, provably
SS Du, W Hu, SM Kakade, JD Lee, Q Lei
arXiv preprint arXiv:2002.09434, 2020
2942020
Algorithmic regularization in learning deep homogeneous models: Layers are automatically balanced
SS Du, W Hu, JD Lee
Advances in neural information processing systems 31, 2018
2572018
An analysis of the t-sne algorithm for data visualization
S Arora, W Hu, PK Kothari
Conference on learning theory, 1455-1462, 2018
2312018
Combinatorial multi-armed bandit with general reward functions
W Chen, W Hu, F Li, J Li, Y Liu, P Lu
Advances in Neural Information Processing Systems 29, 2016
1652016
Provable benefit of orthogonal initialization in optimizing deep linear networks
W Hu, L Xiao, J Pennington
arXiv preprint arXiv:2001.05992, 2020
1492020
Enhanced convolutional neural tangent kernels
Z Li, R Wang, D Yu, SS Du, W Hu, R Salakhutdinov, S Arora
arXiv preprint arXiv:1911.00809, 2019
1452019
Linear convergence of the primal-dual gradient method for convex-concave saddle point problems without strong convexity
SS Du, W Hu
The 22nd International Conference on Artificial Intelligence and Statistics …, 2019
1452019
Width provably matters in optimization for deep linear neural networks
S Du, W Hu
International Conference on Machine Learning, 1655-1664, 2019
1032019
Explaining landscape connectivity of low-cost solutions for multilayer nets
R Kuditipudi, X Wang, H Lee, Y Zhang, Z Li, W Hu, R Ge, S Arora
Advances in neural information processing systems 32, 2019
1002019
Simple and effective regularization methods for training on noisily labeled data with generalization guarantee
W Hu, Z Li, D Yu
arXiv preprint arXiv:1905.11368, 2019
932019
More than a toy: Random matrix models predict how real-world neural representations generalize
A Wei, W Hu, J Steinhardt
International conference on machine learning, 23549-23588, 2022
832022
The surprising simplicity of the early-time learning dynamics of neural networks
W Hu, L Xiao, B Adlam, J Pennington
Advances in Neural Information Processing Systems 33, 17116-17128, 2020
802020
Impact of representation learning in linear bandits
J Yang, W Hu, JD Lee, SS Du
arXiv preprint arXiv:2010.06531, 2020
72*2020
Linear convergence of a frank-wolfe type algorithm over trace-norm balls
Z Allen-Zhu, E Hazan, W Hu, Y Li
Advances in neural information processing systems 30, 2017
672017
How do transformers learn in-context beyond simple functions? a case study on learning with representations
T Guo, W Hu, S Mei, H Wang, C Xiong, S Savarese, Y Bai
arXiv preprint arXiv:2310.10616, 2023
592023
Implicit bias in leaky relu networks trained on high-dimensional data
S Frei, G Vardi, PL Bartlett, N Srebro, W Hu
arXiv preprint arXiv:2210.07082, 2022
592022
Системата не може да изпълни операцията сега. Опитайте отново по-късно.
Статии 1–20