About Me
I am a Ph.D. student in Data Science at the University of Virginia, where I am fortunate to be advised by Dr. Sheng Li at the RISE Lab and Dr. Tom Hartvigsen . My research focuses on developing trustworthy Large Language Models through novel approaches in reasoning mechanisms and uncertainty estimation, leveraging multi-agent systems for enhanced reliability, and investigating human-aligned behaviors with applications in social science and health.
Prior to UVA, I completed my M.S. in Biostatistics at Harvard University, where I worked with Dr. William La Cava on fairness-aware machine learning in healthcare applications. This experience shaped my interest in developing responsible AI systems that can be reliably deployed in critical domains.
Research
Trustworthy Large Language Models (LLM) Reasoning
My research focuses on developing efficient and reliable reasoning mechanisms for LLMs, with particular emphasis on the problems related to efficiency and uncertainty estimation.
Work Under Review
"Bridging causal discovery and large language models: A comprehensive survey."
arXiv:2402.11068 (2024)
A comprehensive survey exploring the intersection of causal discovery and LLMs, outlining future research directions.
"Dynamic Self-Consistency: Leveraging Reasoning Paths for Efficient LLM Sampling."
Introduced RASC framework that systematically reduces LLM Self-Consistency sampling by ~70% while maintaining accuracy by evaluating both reasoning paths and final answers.
"CoT Rerailer: Enhancing the Reliability of Large Language Models in Complex Reasoning Tasks."
A novel two-stage system that first identifies flawed reasoning then fixes errors in LLM outputs using consistency checks and multi-agent debate.
Social Intelligence in LLM Systems
I investigate how LLMs understand and interact with social contexts, focusing on confidence calibration, bias mitigation, and social behavior modeling.
Work In Progress
Fake Confidence and Hallucinations in LLMs
A social-psychological approach to measure and control LLMs' overconfidence through fine-tuning and prompts in specific social scenarios. Developed metrics to evaluate fake confidence and its relationship with hallucinations.
Linguistic Bias in LLMs
Developed systematic testing protocols to compare LLM responses between standardized English and African American English Preference tuned LLM, identifying bias patterns across social scenarios and proposing targeted interventions for fairness improvement.
Applications in Healthcare and Social Science
My work in healthcare AI focuses on developing fair and interpretable machine learning systems, particularly in clinical decision support and mental health applications.
Published Work
"Intrapartum electronic fetal heart rate monitoring to predict acidemia at birth with the use of deep learning."
American Journal of Obstetrics and Gynecology (2024)
"Two-step light gradient boosted model to identify human west nile virus infection risk factor in Chicago."
PLOS ONE 19(1): e0296283 (2024)
"Fair admission risk prediction with proportional multicalibration."
Conference on Health, Inference, and Learning (2023)
Work Under Review
Proactive LLM Doctor
Developed a framework integrating structured domain knowledge with LLMs to create proactive conversational agents for Mental Health Differential Diagnosis.
Mind AI's Mind
Created a medical AI explanation pipeline combining XAI techniques and LLM-based reasoning for transparent depression diagnosis, achieving significant AI-Expert agreement scores and implementing a RAG-enhanced LLM reasoning system for natural language diagnostic reports.
Education
University of Virginia
08/2023 - 05/2027 (Expected)
Doctor of Philosophy in Data Science
Research Focus: LLM Responsible Reasoning & Agents & Social Intelligence
Advisors: Dr. Sheng Li & Dr. Tom Hartvigsen
Harvard University
08/2021 - 05/2023
Master of Science in Biostatistics (Concentration in Data Science)
GPA: 3.89
University of Illinois-Urbana, Champaign
09/2017 - 12/2020
B.S. in Statistics (Summa Cum Laude & James Scholar)
Minor: Computer Science & Mathematics
GPA: 3.94
Teaching Experience
University of Virginia
- CS 5012: Foundations of Computer Science (Summer 2024)
- DS 6310: Statistics Inference Theory II: Inference & Prediction (Spring 2024)
- DS 6030: Statistical Learning (Fall 2023)
Harvard University
- CS 109B: Data Science 2: Advanced Topics in Data Science (Spring 2023)
Academic Service
Conference Reviewer
- Neural Information Processing Systems (NeurIPS) 2024
- International Conference on Learning Representations (ICLR) 2025
- International Conference on Artificial Intelligence and Statistics (AISTATS) 2025
Journal Reviewer
- IEEE Computational Intelligence Magazine (CIM)
Technical Skills
Machine Learning & AI
PyTorch, Hugging Face, LangChain/LangGraph, Transformers, Unsloth, LightGBM
Big Data & Cloud
PySpark, AWS (EC2/SageMaker), Snowflake, Vertica, Dask
Programming
Python, C/C++, R, SQL, NoSQL, Bash
Development Tools
Docker, Git, Unix, VSCode, Cursor