Picture for Fanghui Liu

Fanghui Liu

Scalable Learned Model Soup on a Single GPU: An Efficient Subspace Training Strategy

Add code
Jul 04, 2024
Viaarxiv icon

Benign overfitting in Fixed Dimension via Physics-Informed Learning with Smooth Inductive Bias

Add code
Jun 16, 2024
Viaarxiv icon

High-Dimensional Kernel Methods under Covariate Shift: Data-Dependent Implicit Regularization

Add code
Jun 05, 2024
Viaarxiv icon

Revisiting character-level adversarial attacks

Add code
May 07, 2024
Viaarxiv icon

Learning with Norm Constrained, Over-parameterized, Two-layer Neural Networks

Add code
Apr 29, 2024
Viaarxiv icon

Robust NAS under adversarial training: benchmark, theory, and beyond

Add code
Mar 19, 2024
Viaarxiv icon

Generalization of Scaled Deep ResNets in the Mean-Field Regime

Add code
Mar 14, 2024
Viaarxiv icon

Can overfitted deep neural networks in adversarial training generalize? -- An approximation viewpoint

Add code
Jan 24, 2024
Viaarxiv icon

Efficient local linearity regularization to overcome catastrophic overfitting

Add code
Jan 21, 2024
Viaarxiv icon

On the Convergence of Encoder-only Shallow Transformers

Add code
Nov 02, 2023
Viaarxiv icon