A Letter to Candidates

Dear Prospective Students,

Thanks for your interest in being a part of our group! I have multiple fully-funded Ph.D., MPhil, postdoc, research assistant, and intern positions available at the AI Thrust, HKUST (Guangzhou).

My Goal and Principles

My goal is to conduct impactful research for the benefit of society (AI for social good). I aim to push our research frontiers towards autonomous machine intelligence. Therefore the primary motivation for publishing papers should be to make a significant impact on the research community. I highly value open-source, transparency and collaboration as I believe they are the key factors for the success of our community. As researchers, we should also keep open-minded and be curious about areas outside of our field, so we could continue to grow.

I also believe we are engineers first, and then we are AI research scientists.

"Talk is cheap. Show me the code."

My Operational Style

I hope to build a focused and collaborative group. I think weekly group meetings are valuable to everyone to learn a bit about the research outside of their own area, as long as they are kept short and efficient. Then I would also schedule one-on-one meetings on a need basis. I try to keep normal working hours myself (so please don’t send me messages late at night) and tend to come to the office early. I don’t think long working hours are necessary. If you are focused and continuously work 40-50 hours a week for four years, you are bound to succeed on any research topic. Eventually, I hope my team members can be happy and proud of their own work.

Ongoing Projects

Please refer to this page.

Expectations for Ph.D. Students

Congrats! Please consider this as a significant financial investment (~1M over four years) from the University and I in yourself. My primary goal is to help you grow into the experts in your own field. I believe a successful Ph.D. should have name recognition among fellow researchers by the time of thesis proposal (ideally with >400 citations). In addition, I expect my students to open-source all of their published work whenever it is permitted. I think in the first year, one top-tier publication is good enough. In the coming years, I would like to see growth and push for more continuous and impactful work. Keeping track of progress on all the research work is important as well. Take a read on Prof. Mingming Cheng's advice. In terms of university-level graduation requirement, you only need to propose and defend your thesis successfully, which is the same as CMU.

I want to join HKUST (GZ). What graduate programs should I apply to?

You could apply to the MPhil or Ph.D. program of the AI Thrust.

I am a HKUST (GZ) student. How do I join your group?

Send me an email and/or drop by my office - I'd be happy to chat!

Are you accepting RA/interns/visitors?

Yes!

Should I contact you before applying to HKUST (GZ) for admission?

Admissions across all Ph.D./MPhil programs are done by department-level committees but my endorsement would be a decisive vote for your acceptance. So you could send me the documents listed below if you are interested in my research directions.

Application Documents

I will carefully consider all applications. Due to the large volume of applications, only shortlisted applicants will be contacted within a month.

Advice on Research

Be the first. Be the simplest one. Then be the flashy one.

More about me (and other random thoughts)

I spent six wonderful years with my Ph.D. advisor, Prof. Alex Hauptmann. Alex is one of the most important people in my life and I have learned a great deal from him, both as a good researcher and as a good person. The only complaint I would have with him would be that he did not push me enough during my Ph.D. studies. Take a look at how chilled I was. Therefore I would try to strike a balance between his style and a stricter one. I hope my students can achieve their own level of work-life-balance since different people might have different ratios on how they want to spend their time. The principle, as mentioned before, is that I want my students and I, to be happy first, and then also be proud of their own work. The assumption, of course, is that you and I want to achieve the same goal, which is making an impact in the research community (not just wanting to graduate). I liked Andrej Karpathy's A Survival Guide to a Ph.D..

After my Ph.D. graduation, I received all the big packages from big companies you can think of (Huawei, Tencent, Momenta, Alibaba, Google), and with the top 1% - 10% TC among Ph.D.s. However, I found that I wasn’t making the impact that I was hoping for. And most of the time, it is impossible to dive deep into a research direction without the pressure of making a profit for the company in the short term. Also, I want my students to have a clear sense of credit for their work, especially when collaborating with each other. It is beneficial to communicate early about who is the first author/equal contribution, etc.