Zhenguo Li

4 papers Huawei Noah's Ark Lab, Hong Kong

• FUDOKI: Discrete Flow-based Unified Understanding and Generation via Kinetic-Optimal Velocities
• Masked Diffusion Models as Energy Minimization
• Multi-Objective One-Shot Pruning for Large Language Models
• Speculative Jacobi-Denoising Decoding for Accelerating Autoregressive Text-to-image Generation