Yunwei Zhao

I'm a master student in Computer Science at Cornell University. I am actively seeking PhD opportunities for the fall of 2024!

As a highly self-motivated and independent person, my time at the University of Washington has been marked by significant accomplishments. At University of Washington, my work under Professor Sheng Wang focused on designing robust evaluation protocols of LLMs from human feedback and advancing multilingual and multimodal translation methods for science (Gemini, Crit2SQL, and BioTranslator). This research experience solidified my interest and commitment to adaptive and trustworthy NLP. I have also worked with Professor Richard Anderson on user-centric technology that aids people in developing regions (eKichabi V2). All of these experiences have solidified my interest in user-centric NLP systems, which enhance my passion for the broader field of human-AI interaction.

At Cornell, I have been actively working on research, startup projects, and internships to design user-centric NLP systems in practice and integrate them into real products to study human-LM interaction.

Email  /  CV  /  Research Statement  /  Github  /  Google Scholar  /  Twitter

Research

My research interests lie in adaptability, trustworthiness, and human-AI interaction. I aim to bridge the gap in user engagement by integrating real-time human feedback with AI design to develop AI agents that can learn from human behaviors, communicate naturally, and explain their processes. I am also broadly interested in the intersection of Human-Computer Interaction and Natural Language Processing.

I have two authored papers: one is eKichabi V2 with Professor Richard Anderson where I actively involved in app design and implementation, user testing, data analysis, and manuscript refinement; the other one is Crit2SQL where I contribute to dataset analysis and development with Professor Sheng Wang.

Recent Works

Miscellanea

NAACL
2022

2022