Our Research
We model, analyze, and explain how humans and AI agents behave and make decisions — from individuals to markets.
Below are some selected examples of our recent works.
AI-driven Solutions For Enhancing Social and Mental Well-being
We build agentic systems with human experts in the loop to deliver accessible, interactive, and personalized health management and intervention for users' emotional and social well-being. Through collaboration with medical specialists and institutions, we iteratively optimize the design (including algorithms, functional features, and the user interface) to improve user experience and well-being outcomes.
Computational Science of Human-AI Affective Relationship
A new category of human experience is taking shape. People are forming affective relationships with AI systems (e.g., companions, therapeutic interfaces, and emotionally responsive agents) at a pace and scale that no prior technology has approached. They are ongoing, emotionally textured, sometimes intimate, and increasingly consequential for how people make decisions and regulate their wellbeing. We are devoted to the computational science of human-AI affective relationships. We ask how such relationships form, evolve, and influence human cognition, behavior, well-being, and social life. We treat these relationships as a distinct phenomenon requiring its own theoretical vocabulary, measurement tools, and empirical evidence.
User-centered Design for Professionals
We leverage AI to empower healthcare professionals in navigating caregiving decisions and enabling more effective engagement with patients/consumers. For example, we collaborate with clinicians and hospitals to develop AI-assisted systems that enable more precise triage, follow-up, and intervention planning across diverse clinical presentations.