I am an M.S. student in Computer Science at UC San Diego, advised by Prof. Hao Zhang in the Hao AI Lab, and I hold a B.S. from ShanghaiTech University advised by Prof. Kewei Tu. My research lies at the intersection of Machine Learning Systems and generative models, with a focus on efficient inference, scalable video generation, and world models.

I am currently a Student Researcher at ByteDance Seed Infra, where I own the internal TPU kernel agent system with multi-agent scheduling and a self-evolving knowledge base, accelerated attention end-to-end by 25%, and optimize MoE and fused TPU kernels for large-scale LLM inference. At the Hao AI Lab, I am a core contributor and code owner of FastVideo (3.7K+ stars), leading next-gen video sparse attention research and co-leading quantization-aware distillation releases. I helped build and ship DreamVerse, a real-time video generation workspace that streams 1080p clips on a single GPU.

I am a core contributor to HiLS-Attention, a learned chunk-wise sparse attention achieving 64× context-length extrapolation with up to 13.5× prefill and 15.7× decode speedup — I built its official SGLang serving backend. Earlier, I was the lead author of FlashMHF, a Multi-Head FFN architecture backed by IO-aware Triton/CUDA kernels that cuts peak memory by 3–5×.