Picture for Yu-Xiang Wang

Yu-Xiang Wang

Stable Minima Cannot Overfit in Univariate ReLU Networks: Generalization by Large Step Sizes

Jun 10, 2024
Viaarxiv icon

Neural Collapse Meets Differential Privacy: Curious Behaviors of NoisyGD with Near-perfect Representation Learning

May 16, 2024
Viaarxiv icon

Differentially Private Reinforcement Learning with Self-Play

Apr 11, 2024
Figure 1 for Differentially Private Reinforcement Learning with Self-Play
Viaarxiv icon

CPR: Retrieval Augmented Generation for Copyright Protection

Mar 27, 2024
Figure 1 for CPR: Retrieval Augmented Generation for Copyright Protection
Figure 2 for CPR: Retrieval Augmented Generation for Copyright Protection
Figure 3 for CPR: Retrieval Augmented Generation for Copyright Protection
Figure 4 for CPR: Retrieval Augmented Generation for Copyright Protection
Viaarxiv icon

Privacy Profiles for Private Selection

Feb 09, 2024
Viaarxiv icon

Permute-and-Flip: An optimally robust and watermarkable decoder for LLMs

Add code
Feb 08, 2024
Viaarxiv icon

Weak-to-Strong Jailbreaking on Large Language Models

Add code
Feb 05, 2024
Viaarxiv icon

Online Feature Updates Improve Online (Generalized) Label Shift Adaptation

Feb 05, 2024
Viaarxiv icon

Near-Optimal Reinforcement Learning with Self-Play under Adaptivity Constraints

Feb 02, 2024
Viaarxiv icon

Improving the Privacy and Practicality of Objective Perturbation for Differentially Private Linear Learners

Dec 31, 2023
Viaarxiv icon