Undergraduate Researcher · Zhejiang University
I am a senior undergraduate at Zhejiang University, pursuing a dual degree in Statistics and Information Management & Information Systems. My research focuses on making image and video generation high-quality, efficient, and interactive. I am applying to PhD programs for Fall 2027.
Most recently, I was a research intern at PixAI, working with Prof. Chenxi Liu (University of Victoria) on interactive image generation. Before that, I spent eight months at the ZIP Lab advised by Prof. Bohan Zhuang, working on sparse attention and KV-cache design for interactive, minute-long video generation.
I also enjoy high-performance computing: our team placed 1st globally at the SC25 IndySCC, and I look after the GPU infrastructure at ZJU X-Lab.
Awarded the National Scholarship (top 0.2%) for the third consecutive year.
HimmBO, my first-author work on interactive LoRA merging, is on arXiv.
WorldAttention is on arXiv. It received Accept / Accept / Accept (5/5/5) reviews at NeurIPS 2026.
BIFE is accepted to SIGGRAPH Asia 2026 Technical Communications.
Joined PixAI as a research intern, working with Prof. Chenxi Liu on interactive image generation.
Our team won 1st place globally at the SC25 IndySCC Student Cluster Competition.
Awarded the National Scholarship (top 0.2%) for the second consecutive year.
Joined the ZIP Lab at Zhejiang University, advised by Prof. Bohan Zhuang.
National 2nd Prize at the 12th National Parallel Application Challenge (PAC).
Designed and built the official website of ZJU X-Lab from scratch.
My study notes are online: My Notebook.
1st place in the National Finals of the ICT Competition.
* equal contribution
arXiv
2026 · Preprint
arXiv preprint, 2026
Two-stage preferential Bayesian optimization scales interactive LoRA merging from a few dozen to 1,080 adapters. A style-aware kernel shares preferences across similar adapters, so it needs under 1/10 of the inference used by prior work.
NeurIPS
2026 · Submission
NeurIPS 2026 submission Review scores: 5/5/5 · Accept / Accept / Accept
Attention heads in video diffusion transformers are sparse in very different ways. Hybrid Sparse Attention pairs a linear global branch with head-adaptive block-sparse attention for a 2.21× end-to-end speedup at matched quality.
SIGGRAPH Asia
2026
arXiv
2025 · Preprint
arXiv preprint, 2025
An evolution-driven explorer searches the transfer function space while an MLLM evaluates renderings, forming a trial–insight–replanning loop that turns user intent into volume visualizations.
Efficient pixel-space diffusion · Advised by Prof. Sheng Li
Interactive image generation · Advised by Prof. Chenxi Liu (University of Victoria)
Efficient video generation & world models · Advised by Prof. Bohan Zhuang
Database agent framework · Zhejiang University
Student cluster competitions & HPC application optimization
B.S. in Statistics & B.Mgt. in Information Management and Information Systems (dual degree) · Hangzhou, China
GPA 4.19/4.3 (92.64/100)Rank 1/20
Student-run innovation organization
Algorithm development · Hangzhou