Undergraduate Project 2025

I mentor a practical course for senior BME undergraduate students. This semester, we will investigate the use of LLMs in medical applications.

Description

The students are grouped into three teams, each focusing on a critical aspect of medical LLM applications:

Project Topics

1. 🩺 Hallucination Detection - Developing methods to identify and mitigate false information in medical LLM outputs
2. 🛡️ AI Safety - Ensuring responsible and safe deployment of LLMs in healthcare settings
   
3. 🧠 Reasoning - Enhancing logical reasoning capabilities of medical LLMs for better clinical decision support

We will provide foundational code for each project to help students get started, and expect them to conduct research and report their findings.

Project Materials

The example projects can be previewed below:

🩺 Hallucination Project 🛡️ Safety Project 🧠 Reasoning Project

Note

Download Materials: The complete project code and datasets can be downloaded from Baidu Yun using password xd3w.

This course takes place in Room B5011, No. 2 Interdisciplinary Research Building on Tuesday afternoons (15:00–16:00) during the first semester of the 2025–2026 academic year on the following dates.

Schedule

Week	Date	Plan
4	09/30/2025	Course Introduction
8	10/28/2025	Stage Discussion I
12	11/25/2025	Stage Discussion II
16	12/23/2025	Student Presentation

Recommended Reading Materials

• ChatGPT Prompt Engineering for Developers.
• Building Large Language Models (LLMs).