About

I have a formal background in mathematics, science, design and engineering, and have been working with machine learning system (e.g., Computer Vision, Natural Language Processing, Reinforcement Learning) in large-scale distributed settings with love of GNU, Linux and Emacs for more than 10 years. I am also perhaps knowledgeable in most disciplines, e.g., physics, history, economics, or philosophy—for the liberal arts and humanities part, refer to this page.

I got a Bachelor degree in Computer Science from the Special School for Gifted Young at University of Science and Technology of China in 2014, and a Master of Philosophy degree in Electronic Engineering from Chinese University of Hong Kong in 2017. I got trained in innovation and design at Global Engineering Design Innovation program at Stanford University, and Entrepreneurship and Maker Skills Integrator program at Massachusetts Institute of Technology.

I experienced academia by interning at Microsoft Research Asia and working as an associate researcher at South China University of Technology, at which the responsibilities include mentoring PhD candidates, and publishing at top-tier avenues (e.g., TPAMI) on topics such as Deep Learning and Statistical Learning. I also spent two years as an independent researcher to work out a science of Deep Neural Networks (speaking from the philosophy of science). Those were my basic (blue skies) research days in which I clarified the framework to understand (artificial) intelligence.

After the framework has been set up, the work becomes more and more applied and technological, or in other words, more and more main stream. Thus, I am currently work in the capacity of staff R&D engineer at Baidu Inc.. More specifically, I am working under the theme of embodied intelligence, specifically focusing on the alignment of large language model for the time being—details given on the page Research & Development.