Reasoning
How can we understand reasoning structure in language and multimodal models—what makes their reasoning successful, and when does it fail?
Chenrui Fan · College Park, MD
Ph.D. Student in Computer Science
Hi, I’m Chenrui.
I use interpretability to study reasoning capabilities and agent behavior in language and multimodal foundation models—how they work, when they fail, and how they can be evaluated, steered, and improved.
I am a Ph.D. student at the University of Maryland, advised by Prof. Soheil Feizi. My research sits at the intersection of language model reasoning, interpretability, and reliable agent decision-making.
I am actively looking for Summer 2027 internship opportunities.
Research
Research interests
My research spans three closely connected fields.
Interpretability
How can we understand what foundation models learn and how they make decisions, then turn those insights into actionable interventions?
Agent
How can agents make reliable decisions, learn continually from experience, and safely evolve as tasks and environments change?
Contact
Interested in working together?
I’m happy to talk about research, collaboration, or new ideas.