Rui Wang

Computer Science & Artificial Intelligence

Undergraduate researcher at HKUST exploring Large Language Models, decision-making under uncertainty, and confidence calibration in AI systems.

Contact Me Google Scholar

About Me

I am a fourth-year undergraduate student at The Hong Kong University of Science and Technology (HKUST), pursuing a Bachelor of Engineering in Computer Science with an Extended Major in Artificial Intelligence.

My research focuses on Large Language Models (LLMs), particularly in areas such as decision-making under epistemic uncertainty and confidence calibration in Retrieval-Augmented Generation (RAG) systems. I am passionate about understanding and improving the reliability of AI systems.

Currently, I am on a semester exchange at The University of Texas at Austin, where I am taking advanced courses in Software Engineering and Machine Learning for Edge AI.

3.98

Cumulative GPA

4.07

Major CGA

Papers Under Review

Education

The Hong Kong University of Science and Technology

Sep 2023 - May 2027 (Expected)

Bachelor of Engineering in Computer Science - Extended Major in AI

Cumulative GPA: 3.983 / 4.30
Major CGA: 4.073 / 4.30
Honors: Dean's List (All Semesters)
Scholarship: University Continuing Scholarship (Total 60,000 HKD)

The University of Texas at Austin

Jan 2026 - May 2026

Semester Exchange, Electronic and Computer Engineering

Coursework: Software Engineering, Machine Learning & Data Analytics for Edge AI

Research Experience

Under Review (ACL) Aug 2025

Revealing Instability of Decision-Making under Epistemic Uncertainty

First Author

Designed a three-stage experiment to evaluate LLM decision-making, benchmarking model behaviors against Prospect Theory axioms such as Loss Aversion and Asymmetric Risk Preferences
Revealed that modeling these systems with economic frameworks is not consistently reliable, particularly under diverse linguistic forms, highlighting potential Model Risk and instability

arXiv Paper

Under Review (ACL) Jan 2026

Noise-Aware Verbal Confidence Calibration for LLMs in RAG Systems

Co-first Author

Developed NAACL, a framework that resolves RAG overconfidence by training models to explicitly discern retrieval noise and calibrate their verbal confidence

arXiv Paper

Technical Skills

Machine Learning & Programming

Python PyTorch NumPy Scikit-Learn C++

Software Development

Git Bash Scripting GitHub Copilot Cursor

LLM Operations

vLLM LlamaFactory

Relevant Coursework

Data Structures (A+) Knowledge Discovery (A) Operating Systems Algorithms Probability (A+) Linear Algebra (A+)

Leadership & Activities

Teaching Assistant

COMP1021: Intro to Computer Science

Designed coding assignments and conducted tutorials for 50+ first-year undergraduates; received positive feedback for clarifying complex CS concepts

Mentor

CSE Peer Mentor Program

Provided academic guidance and career planning advice to junior students, fostering a collaborative learning environment

Languages

English (IELTS 7.5), Mandarin (Native), Cantonese (Passive)

Interests

Strategic Games, Poker, Badminton

Get In Touch

I'm always interested in discussing research opportunities, collaborations, or just connecting with fellow researchers and students. Feel free to reach out!