Zhiyu An
I'm a Postdoctoral Scholar at UC Merced. My research applies reinforcement learning to LLMs, civil, and agricultural systems, enabling safe AI-of-Things (AIoT). I focus on building systems that can anticipate uncertainties, act robustly in rare scenarios, and align to user preferences and intents.
Core research themes include:
Uncertainty quantification and intrinsic curiosity in learning systems;
Robustness against rare events;
Alignment and steering to human preferences and intents.
You can find my work in top venues across machine learning, design automation, and IoT including ICLR, AAAI, DAC, KDD, L4DC, BuildSys, and IEEE IoT-J. I received Best Paper Award Runner-Up at BuildSys 2023 and Best Paper Award at HICSS 2025.
I received my Ph.D. from UC Merced, where I was supervised by Wan Du and worked on AI-enabled smart buildings and smart agriculture to help saving our valuable resources and building climate resiliency.