Chenxi (Chelsea) Yuan
I build reliable AI that learns and makes decisions under uncertainty, incomplete information, and heterogeneous real-world data — with a focus on healthcare.
Background
Research
AI methods that stay reliable when data are heterogeneous, incomplete, and high-stakes.
Uncertainty & safety of LLMs
Large language models and agents that know when they don't know, with calibrated uncertainty and safety guardrails.
LLM agents that reason & learn
Agents that retrieve and reason over structured knowledge, learn from experience, and guide conversations with people.
Fair & private medical AI
Verifying that clinical models treat patient groups equitably and that generative models protect patient privacy.
Generative AI for Alzheimer's
Diffusion models and multimodal fusion that recover, harmonize, and interpret neuroimaging for Alzheimer's disease.
News
Ph.D. students and interns in generative AI
I'm looking for highly self-motivated Ph.D. students and interns. Email your CV, transcript, and anything else you'd like to share.
GRE is required. Please apply only if you have taken the GRE or plan to take it soon.
My lab is supported by
