CV

Education

• The Hong Kong University of Science and Technology (HKUST)
  • B.Sc. in Computer Science and Mathematics (Double Major)
  • Sep. 2023 – June 2027 (Expected)
  • Cumulative GPA: 3.923 / 4.3
  • Selected Coursework: Advanced Deep Learning Architecture (Postgrad level), Graph Neural Networks, Large Language Models, Honors Probability.
• École Polytechnique Fédérale de Lausanne (EPFL)
  • Exchange Student
  • Spring 2026 (Nominated)
• Stanford University
  • International Honors Program (Summer Session)
  • June 2025 – Aug. 2025
  • Coursework: Machine Learning (CS229), Design and Analysis of Algorithms (CS161)

Research Experience

• Pathology Report Generation via Multi-modal Learning (Sep. 2025 – Present)
  • Research Assistant, Supervised by Prof. Hao Chen, HKUST Smart Lab
  • Proposed a novel multi-modal framework to automatically generate pathology reports from Whole Slide Images (WSI).
  • Developing a two-stage training pipeline for pathology report generation, leveraging a highly structured, high-quality dataset to enhance generation accuracy.
  • Addressed the scarcity of high-quality paired WSI-report datasets by constructing a large-scale instruction-tuning dataset.
  • Designed a modality-alignment algorithm to enhance consistency between visual features and text descriptions, and proposed a disentanglement method to decouple template artifacts from diagnostic information.
  • Leveraged State-of-the-Art Vision-Language Models (VLMs) to assist pathologists, targeting measurable improvements in clinical workflow efficiency.
• Cancer Diagnosis and Prognosis with Whole Slide Images (Sep. 2024 – Sep. 2025)
  • Undergraduate Researcher, Supervised by Prof. Hao Chen (UROP), HKUST Smart Lab
  • Conducted a comprehensive survey on deep learning methods for cancer diagnosis, synthesizing insights from over 10 seminal papers and recent benchmarks.
  • Implemented and evaluated deep learning baselines, including MotCAT and MACT, to analyze performance gaps in cancer subtype classification.
  • Identified key computational bottlenecks inherent in gigapixel resolution and high-dimensional feature spaces that limit traditional ML models.
• iFLYTEK Co., Ltd. (June 2024 – Aug. 2024)
  • AI Data Processing Intern, Hefei, China
  • Developed automated Python scripts for large-scale web data collection, improving data acquisition efficiency.
  • Preprocessed unstructured data to support the unsupervised training of Large Language Models (LLMs).
  • Analyzed data quality and summarized findings in technical reports presented to the senior research team.

Projects

• Zero-Shot Spatial Reasoning with Large Language Models (Mar. 2025 – May 2025)
  • Course Project
  • Investigated the spatial reasoning capabilities of LLMs in zero-shot settings using custom prompt engineering.
  • Designed and executed experiments to evaluate model performance on complex geometric reasoning tasks.

Awards & Honors

• Dean’s List, HKUST (2023 – 2025, Awarded 4 times)
• University Scholarship (Top 2% Students), HKUST (2023 – 2024)

Skills

• Languages: Mandarin (Native), English (Professional Proficiency, TOEFL: 102)
• Programming: Python (PyTorch, TensorFlow), C++, MATLAB, R, HTML/CSS
• Tools: Linux, Git, Docker, LaTeX, Vim

Publications