About me

Hello, I am a Computer Science PhD student at Johns Hopkins University Center for Language and Speech Processing (CLSP) working on Natural Language Processing, advised by Prof. Mark Dredze. Before joining JHU, I was a student at Yale University, where I first learned and researched NLP under the guidance of Prof. Arman Cohan. My research interests are natural language processing, and applied mathematics and statistics, especially in health, social, and business applications. I regard many practical problems as an optimization problem with partially observed data. How to make the most use of the limited data and how to maximize the quantified outcome through simple, elegant ideas and rigorous definitions are my primary focuses during research.

My current research interests include (take the medical domain for example):

  • Evaluation of Open-ended Machine Generated Text: Past evaluations require human-written answers as a gold standard, and compare lexical or semantic similarities of machine-generated text with gold answers. How can we evaluate diverse LLM-generated text with no human annotations? My current work focuses on factuality, completeness, and safety evaluation of LLM answers to patient questions.

  • Utilization of Web and Knowledge: Web information is mixed with misinformation and factual information. How can we identify misinformation based on the goal of study and optimally utilize the filtered knowledge to reach our goal?

  • Evaluation of Agents and Understanding Model Behaviors

To scale up affordable and accessible healthcare services, I envision one day chatbots can act at a trained clinical expert level to understand, comfort, and guide patients to recover.

If you are a student looking for research opportunities in clinical/medical NLP, please fill out this form and email me with your research interest and NLP-related experience.

Publications

MedScore: Generalizable Factuality Evaluation of Free-Form Medical Answers by Domain-adapted Claim Decomposition and Verification

Heyuan Huang, Alexandra DeLucia, Vijay Murari Tiyyala, Mark Dredze

Average Rating at Top 35% of accepted papers at COLM 2025 anonymous review period (Top 11% of all submissions) / code

Medical Text Simplification: Optimizing for Readability with Unlikelihood Training and Reranked Beam Search Decoding

Lorenzo Jaime Yu Flores, Heyuan Huang, Kejian Shi, Sophie Chheang, and Arman Cohan

EMNLP 2023 Findings / code

MedExpert: An Expert-Annotated Dataset for Medical Chatbot Evaluation

Mahsa Yarmohammadi, Alexandra DeLucia, Lillian C Chen, Leslie Miller, Heyuan Huang, …, Mark Dredze

Machine Learning for Health 2025 / code

Service

Conference Reviewer: COLM 2026, ACL 2024

Committee: CLPsych 2025

Teaching

Head Teaching Assistant in Computer Science 601.486/686 Artificial Intelligence System Design and Development, Johns Hopkins University, Aug/2025 - Dec/2025

Teaching Assistant in CPSC 477/577 Natural Language Processing, Yale University, Jan/2024 - May/2024

Miscellaneous

I love to help people take photos, and my aesthetic composition always gives them happiness and touching feelings during traveling and daily life.

My parents raised a dog during my undergraduate time. She is 5 years old now and has a similar personality to mine when I was a kid… 🤨

My favorite book and its adapted musical is Les Misérables by Victor Hugo.