A research journey,
across disciplines.

I am a Research Scientist at the London Research Center of Huawei Noah’s Ark Lab, where I collaborate with Prof. Jun Wang. I build AI methods and scientific software to understand and discover materials. My research connects scientific agents, crystal representation learning, and Bayesian optimization, with the goal of turning experimental data into physical insight and better experimental decisions.

I led the development of Ganjiang (干将), a self-learning scientific agent for X-ray diffraction. Ganjiang brings phase identification, multiphase decomposition, and structure refinement into a unified workflow, and turns validated analytical experience into reusable skills. This work advances my broader aim: scientific AI that reasons with physical evidence and improves through experience. Read the paper

My work spans the diffraction analysis pipeline: large-scale simulation with SimXRD (ICLR 2025), phase identification with XMatcher and XQueryer (NSR), multiphase analysis with XDecomposer (NeurIPS 2026), and physics-constrained refinement with WPEM. I also developed PRDNet (ICLR 2026) for crystal property prediction and Bgolearn for Bayesian experimental design. Bgolearn was selected for support under the Shanghai Municipal Commission of Economy and Informatization’s Open Source Project Program.

I received my Ph.D. from HKUST (Guangzhou), advised by Prof. Tong-Yi Zhang, and visited City University of Hong Kong to work with Prof. Yang Ren. My background combines mechanics at Shanghai University and chemical machinery at Beijing University of Chemical Technology, with research experience at Shanghai AI Lab, GreenDynamics, AI Lab, The Yangtze River Delta, and Zhejiang Lab.

I review for ICLR, NeurIPS, ICML, AAAI, KDD, and journals including Communications Materials, Journal of Applied Crystallography, EAAI, Results in Engineering, and MGE Advances. Outside research, I enjoy running and strength training.