About me

Hi! I’m a 2nd-year Master’s student in Electrical and Computer Engineering (Machine Learning and Data Science Track) at the University of Southern California.

I work with Salman Avestimehr in the vITAL Lab, where I study uncertainty quantification in LLMs for reliable decision making. I also collaborate with Sunwoo Lee, focusing on parameter-efficient fine-tuning and communication-efficient federated learning. Previously, I received my B.S. in Electronic Engineering from Sogang University, where I worked on federated learning and distributed optimization with Hongseok Kim.

📰 News

• May 16, 2025 – Our paper “Reconsidering LLM Uncertainty Estimation Methods in the Wild” was accepted to ACL 2025! I’ll be presenting it in Vienna — see you there! 🎉
• May 05, 2025 – I attended the ATHENA AI Summit 2025 — “Big AI for Small Devices” — at Duke University. It was a great opportunity to connect with distinguished faculty members and leading researchers in the field!
• May 02, 2025 – Our library for assessing the truthfulness of LLMs, TruthTorchLM, officially released!
• May 02, 2025 – I attended the USC-Amazon Center Annual Symposium at USC. It was inspiring to explore recent advances in AI agents, reasoning, and trustworthy machine learning.
• Aug 26, 2024 – I started my M.S. in Electrical and Computer Engineering at USC. Excited for the journey ahead!
• May 22, 2024 - I gave a talk at Sogang University about Federated Learning using TCP socket communication.
• Feb 20, 2024 – I graduated from Sogang University with a B.S. in Electronic Engineering!

🎓 Education

University of Southern California (USC), Los Angeles, USA
M.S. in Electrical and Computer Engineering
Aug. 2024 – May. 2026 (Expected)
GPA: 4.00 / 4.00 – MS Honors Fellow

Sogang University, Seoul, Republic of Korea
B.S. in Electronic Engineering, Micro-degree in Artificial Intelligence
Mar. 2018 – Feb. 2024
GPA: 3.91 / 4.3 – Magna Cum Laude

📬 Contact

Feel free to contact me at kangsung@usc.edu or connect via LinkedIn