I am a Research Scientist at Meta AI, where I work on efficient graph learning models for web-scale social recommendation. Prior to Meta, I obtained my Ph.D. in Computer Engineering at University of Southern California, advised by Prof. Viktor Prasanna. My thesis focused on improving the scalability, accuracy and efficiency of large scale Graph Neural Networks. I designed new neural network models, learning algorithms as well as hardware systems for GNN training and inference. My work has led to publications in top venues in both AI (ICLR, NeurIPS, etc.) and systems (FPGA, VLDB, TRETS, etc.). I have received outstanding reviewer awards from ICLR and ICML.
I am broadly interested in problems in ML and system design. Please reach out to me via my personal email if you’d like to discuss.
PhD in Computer Engineering, 2022
University of Southern California
Bachelor of Engineering, 2016
University of Hong Kong
Responsibilities include:
Responsibilities include:
Achievements:
Old GNN models + New data perspective = Surpassing 1-WL, Avoiding oversmoothing & Overcoming neighborhood explosion
Official code: https://github.com/facebookresearch/shaDow_GNN
PyTorch Geometric implementation: https://pytorch-geometric.readthedocs.io/en/latest/modules/loader.html#torch_geometric.loader.ShaDowKHopSampler
Deep Graph Library implementation: https://docs.dgl.ai/en/latest/_modules/dgl/dataloading/shadow.html