Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Background
Professor at the Institute of AI Industry Research (AIR), Tsinghua University
Research interests include Edge AI, large foundation model algorithms, AI inference systems, and novel AI accelerators
Her research enables deployment of complex deep neural networks and large language models directly on consumer devices such as smartphones and PCs, significantly reducing cloud operational costs
Her innovations have been integrated into widely used products including Microsoft Office, Windows, Bing, and Huawei HarmonyOS
Notable contributions include low-bit and table-lookup techniques for LLM inference, large model accelerators using commercial DRAM, StreamMind (a VLM for continuous video understanding), the near real-time GUI agent V-droid, and the Cloud for Science paradigm