MMODELYST
People/Yuzhang Shang
PPL

Yuzhang Shang

Papers (3)
TIDE: Efficient and Lossless MoE Diffusion LLM Inference with I/O-aware Expert OffloadAttend Locally, Remember Linearly: Linear Attention as Cross-Frame Memory for Autoregressive Video DiffusionGood Agentic Friends Do Not Just Give Verbal Advice: They Can Update Your Weights