Q&A with faculty mentor Ben Zhou

Ben Zhou is an assistant professor of computer science and engineering in the School of Computing and Augmented Intelligence. Zhou’s research focuses on controllable and trustworthy reasoning in natural language processing and artificial intelligence, or AI, systems. He began mentoring students in the FURI program in the spring 2025 semester.

What made you want to get involved as a FURI, MORE and GCSP research stipend program faculty mentor?

I was approached by motivated and passionate undergraduate students who have visions and strong research interests. They introduced me to the FURI program, which I believe to be a great opportunity to give undergraduate students a goal and motivation to work toward.

I was an undergraduate student when I first started research in the direction that eventually led to my current position, and I am forever grateful to my faculty advisor who was willing to spend the time and effort to help me as an undergrad to start early. As a result, I deeply recognize how much of a difference it would make for faculty members to advise motivated undergraduate students, and I am willing to contribute my part.

What is your favorite part about seeing your students conduct research?

My favorite part is when students gradually develop unique and novel perspectives and ideas based on the initial feedback, knowledge and pointers I provide. At the same time, it is awesome to see that the undergraduate students I advise eventually pass on their passion and what they learned through outreach activities, such as AI clubs, to eventually motivate more students.

How have your FURI, MORE and GCSP research stipend students had an effect on your research? Have they come up with any research surprises or proposed new directions for your lab?

The research stipend has enabled undergraduate students in the FURI program to develop a self-sustained ability to play with large language models and AI models on their own, without having to compete for resources with PhD level research and making them feel bad about it. This has enabled many opportunities where the students can develop new ideas based on toggling and probing models, and it eventually led to the current research project we are working on.

What have you gained from being a FURI, MORE and GCSP research stipend mentor? How has the experience been rewarding for you?

I have gained a lot, ranging from how to communicate, motivate and best help undergraduate students who may not have a lot of technical experience, but very strong ideologies and motivations, to how to manage and design undergrad-level projects that will eventually lead to top-tier conference publications. This process is rewarding because from a research perspective, I have learned a lot of fresh ideas; from a mentor perspective, I have learned to communicate more effectively; and from a long-term perspective, the students may eventually pursue career paths because of their early experiences, eventually contributing more broadly to their fields.

What advice would you give to students who might be interested in participating in FURI, MORE and the GCSP research stipend program?

I encourage more students to get involved; there are many more things that you can accomplish by starting early, regardless of what background you have. If you are motivated and passionate about something, you should try to make it happen! You will be surprised to see how much you can accomplish.

Why should other faculty members become FURI, MORE and GCSP research stipend mentors?

It is extremely fun to work with self-motivated undergraduate students who are able to self-support the cost of a research project because of the FURI stipend, especially in the AI and computer science fields. Undergraduate students can provide perspectives that are often missing in doctorate-level research, which is sometimes overwhelmed by “standardized approaches” from others’ works. Also, it is very rewarding to work with and mentor undergraduate students, knowing that they could one day contribute to academia because of our early guidance.