 Excess Risk of Two-Layer ReLU Neural Networks in Teacher-Student
Settings and its Superiority to Kernel MethodsInternational Conference on Learning Representations (ICLR), 2022 |
 On Learnability via Gradient Method for Two-Layer ReLU Neural Networks
in Teacher-Student SettingInternational Conference on Machine Learning (ICML), 2021 |
 Non-asymptotic approximations of neural networks by Gaussian processesAnnual Conference Computational Learning Theory (COLT), 2021 |
 Mathematical Models of Overparameterized Neural NetworksProceedings of the IEEE (Proc. IEEE), 2020 |
 Efficient constrained sampling via the mirror-Langevin algorithmNeural Information Processing Systems (NeurIPS), 2020 |
 Generalization bound of globally optimal non-convex neural network
training: Transportation map estimation by infinite dimensional Langevin
dynamicsNeural Information Processing Systems (NeurIPS), 2020 |
 Predicting the outputs of finite deep neural networks trained with noisy
gradientsPhysical Review E (PRE), 2020 |
