Experience

Led informal peer-assisted learning sessions for first-year Mathematics students, helping them discuss core course questions in a friendly and supportive small-group environment. Through this role, I developed my ability to explain abstract mathematical ideas clearly, guide discussion without simply giving away solutions, and coordinate preparation with other student leaders. The experience reflects my strong interest in mathematical teaching, peer mentoring, and building an inclusive academic community.

Represented Mathematics students by collecting feedback on teaching, assessment, workload, learning resources, and the wider academic experience. I helped communicate recurring concerns to staff through the student representation system and Student-Staff Liaison Committee discussions, while also helping close the feedback loop by sharing responses and updates with classmates. This role strengthened my communication, negotiation, and problem-solving skills, especially in turning informal student concerns into constructive and actionable feedback.

Completed at least 25 hours of volunteering with Edinburgh University Students' Association and received the Saltire Awards 'The Approach' certificate in recognition of my contribution to student life and the wider community. This experience broadened my involvement beyond academic study and reflects a sustained commitment to volunteering, community engagement, and supporting student activities.

Served as a volunteer at Yunnan Nationalities Museum and received the title of 'Outstanding Volunteer' for excellent performance during the museum's volunteer service activities. This experience involved public-facing cultural service in a museum setting, supporting the communication of Yunnan's diverse ethnic cultural heritage to visitors. It helped me develop a stronger sense of responsibility, cultural awareness, and confidence in communicating with people from different backgrounds.

Investigated random-walk-based approaches to graph neural networks, focusing on limitations of standard message-passing architectures such as oversmoothing and oversquashing. I explored how random-walk schemes could improve expressivity and studied possible extensions to hypergraphs, with potential applications in biochemical discovery, network analysis, and recommendation systems. This research experience strengthened my background in graph theory, machine learning, and computational experimentation.