Sungmin Kang is a machine learning researcher and engineer with a decade of hands-on experience, currently pursuing an MS at USC and researching uncertainty quantification and truthfulness in LLMs at the vITAL lab. He has designed communication-efficient and parameter-efficient federated/LLM training algorithms that cut communication by up to 83% and fine-tuned models by updating only 0.1% of parameters. His open-source work includes a library of 30 truth evaluation methods for language model generations and practical FL systems deployed on edge devices like Raspberry Pis. Comfortable bridging theory and systems, he has applied transfer learning to multimodal medical prediction and delivered lectures and projects on federated learning. Having served as a sergeant in the ROK Army, he brings disciplined leadership to collaborative research teams in Los Angeles.
10 years of coding experience
1 year of employment as a software developer
Master of Science - MS, Electrical and Computer Engineering (Machine Learning and Data Science), Master of Science - MS, Electrical and Computer Engineering (Machine Learning and Data Science) at University of Southern California
Bachelor of Science - BS, Electronic Engineering, Bachelor of Science - BS, Electronic Engineering at 서강대학교
Open-science repo for our experimental results of automatic software repair on the Defects4J benchmark of Java bugs
Contributions:1 push, 456 branches in 8 months
sciencebenchmarkdefects4jopen-sciencerepair
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.