Cheng-zhong Xu

Towards a Comprehensive Scaling Law of Mixture-of-Experts

Theoretical Analysis of Relative Errors in Gradient Computations for Adversarial Attacks with CE Loss

RLGF: Reinforcement Learning with Geometric Feedback for Autonomous Driving Video Generation

Scaling Laws for Floating–Point Quantization Training

Honey: Harmonizing Progressive Federated Learning via Elastic Synergy across Different Training Blocks