Dang Nguyen

I’m a Computer Science PhD student in the Chicago Human+AI Lab at the University of Chicago advised by Chenhao Tan.

I’m interested in AI evaluation, including how to evaluate the evaluators when using LLMs as judges. What biases and tendencies do models have, and how might their widespread adoption affect human systems, such as in science and communication? When unintended biases get amplified into harmful behaviors via large-scale model-to-model and model-to-human interactions, how can we intervene effectively?

An important version of this problem I have been actively working on is preventing the AI-mediated surge in research output from negatively affecting the quality of conference reviewing. I helped develop OpenAIReview as a potential solution to this issue. I’m also interested in using AI to assist with verification and understanding in mathematics in response to the research acceleration driven by AI.

Outside of research, I’m quite active and you can find me training Brazilian Jiu Jitsu, boxing, or lifting weights. If you’d like to chat, feel free to reach out!