Ze-Wei Liou

Ze-Wei Liou is a second-year PhD student in the Dao AI Lab at Princeton University, advised by Prof. Tri Dao. He received his bachelor's degree in electrical engineering from National Taiwan University. His research focuses on efficient algorithms and systems for LLM inference, with a particular interest in linear attention and hybrid models.
Ze-Wei Liou

Posts by Ze-Wei Liou

Agentic AI / Generative AI

ReplaySSM: Rethinking State Storage for Faster Hybrid-Model Decoding

Hybrid models interleave state-space model (SSM) layers with a few attention layers. NVIDIA Nemotron 3, Qwen3.5 through Qwen3.8, Kimi-K3, and GLM-5.3-Flash are... 14 MIN READ