Bingrui Li

Bingrui's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

authored a paper 2 months ago

On the Optimization and Generalization of Two-layer Transformers with Sign Gradient Descent

Paper • 2410.04870 • Published Oct 7, 2024

authored 2 papers 3 months ago

Memory Efficient Optimizers with 4-bit States

Paper • 2309.01507 • Published Sep 4, 2023

Efficient Hyperparameter Tuning via Trajectory Invariance Principle

Paper • 2509.25049 • Published Sep 29, 2025 • 4