About Me

Hi! I’m YUAN Mei (袁 玫)! I am currently a visiting student at Carnegie Mellon University, working on agentic multimodal systems and embodied AI for automated science and biomedicine.

I received my Master’s degree in Computational Linguistic Engineering from Peking University, where I was fortunately supervised by Prof. LI Boting and Prof. SU Qi. Before that, I obtained my Bachelor’s degree in Information System.

I have interned at Microsoft Research Asia (MSRA), with a focus on time series foundation models, and at Apple Inc., where I led pilot LLM application projects for the sales sector. Additionally, I co-founded TextPanda, an intelligent writing platform featuring our Chinese Spelling Check model ‘Zous-1B’.

My research interest focuses on (Multimodal) Large Language Models and their scientific and creative applications. Specifically:

• Automated scientific discovery.

  Building human-like agentic systems that autonomously monitor industrial and laboratory processes-from biochemical synthesis to manufacturing quality control. These systems perceive multisensory data, analyze temporal dynamics, forecast future states, detect anomalies, and make autonomous decisions with minimal human intervention.

• Digital learning and humanities.

  Designing adaptive tutoring systems with specialized support for human-in-the-loop interactive paradigms to assist personalized education and human learning. Likewise, creating unified frameworks for free-form coherent, controllable storytelling that enable creative expression with global context understanding.

• Unified 3D/4D world models.

  Developing unified, generalist MLLMs that synchonize perception, reasoning, and generation across omni-modalities and dimensions with spatio-temporal awareness—learning from continual sensory experience to understand and create 3D/4D worlds.

My email is meiyuan [at] andrew [dot] cmu [dot] edu or yuanapr97 [at] gmail [dot] com.

---