Xianyu Chen Earns Distinguished Dissertation Award
Department of Computer Science & Engineering recent PhD graduate Xianyu Chen is the recipient of the University of Minnesota’s Distinguished Dissertation Award for Mathematics, Physical Sciences, and Engineering. The honor recognizes outstanding dissertations that represent original work and make a significant contribution to their field. The award includes a $1,000 prize and a nomination for the 2026 Council of Graduate Students/ProQuest Distinguished Dissertation Award.
“Receiving this award is a tremendous honor,” Chen said. “There are so many opportunities for Artificial Intelligence (AI) systems to improve people’s quality of life and well-being, which is the direction I hope to continue pursuing. This recognition gives me confidence to keep walking down that path, and that means a great deal to me.”
Chen’s dissertation, “Building Human-like Machine Intelligence: Advancing Attention by Modeling, Alignment, and Explainability,” explores visual attention as a bridge between perception and reasoning in artificial intelligence systems.
His research looks at how people move their eyes when answering visual questions, enabling AI systems to learn how humans prioritize task-relevant information. He also created a large-scale dataset of real-world how-to scenarios, such as assembling furniture or cooking a recipe. In the dataset, the images are paired with language and include complex steps, helping the AI system to generate step-by-step visual solutions. Additionally, Chen developed a system that predicts where a person would look and provides short explanations for each instance, linking decisions to evidence.
All of this work, Chen explained, aims to make AI more interpretable, transparent, and trustworthy by grounding decisions in human-like visual reasoning. Such advances, he noted, could help researchers detect failures earlier, improve model design, and support applications such as diagnostic tools that show clinicians why an abnormality was flagged or self-driving systems that provide verifiable justification for their choices.
Reflecting on the award, Chen acknowledged the many people who supported his work.
“Professor Catherine Zhao is the best advisor I could have asked for,” Chen said. “She saw something in me before I saw it in myself. She discussed and developed research ideas with me and taught me how to pursue original, rigorous AI research. I feel very lucky to have learned from her as both a scientist and person. I am also grateful to my department, my collaborators, and my parents who have all shaped me into who I am today.”
Learn more about Chen’s work at his personal website.