I am Yi Cao,
a PhD researcher building trustworthy, physics-grounded AI
for materials discovery at Johns Hopkins University.

My long-term goal is to bridge the gap between computational simulations and experimental materials science, enabling a closed-loop design process. With prior wet-lab training in biomaterials, I’ve seen how tedious trial-and-error methods are.

My vision is to build systems that minimize experiments by learning from past data and simulations—so materials discovery becomes faster, deeper, and smarter.

I aim to merge simulation data and historical experiments using advanced ML techniques like active learning and transfer learning to uncover hidden patterns. This allows us to optimize material design with fewer experiments, while gaining more knowledge—accelerating understanding of atomic-level interactions and enabling better materials in fewer cycles.

To me, ML is not magic—it’s a dynamic tool that gains strength when guided by domain knowledge. With strong grounding in materials science, I see ML as a transparent, explainable collaborator. CPUs and GPUs are extensions of human thought. AI and humans co-evolve, inspiring each other.

As Marie Curie once said, "Nothing in life is to be feared, it is only to be understood." Through interdisciplinary research in ML and materials, I hope to help people understand—and therefore face—the world with greater confidence and curiosity.