Summary
Mingxue Xu is a founder and researcher blending entrepreneurship with deep technical expertise in generative language model compression. Currently pursuing a PhD at Imperial College London, Mingxue applies algebraic approaches to make large generative models more efficient, building on prior engineering and product experience at Baidu’s PaddlePaddle team where they co-authored an IEEE conference paper. With four years of industry experience and a strong academic foundation (MEng from USTC and top-ranked BE from China University of Petroleum, Beijing), they bridge research, product, and production deployment. Based in Stony Stratford, Mingxue brings a pragmatic, programmer’s mindset—self-described as “a little happy programmer”—and a founder’s drive to turn cutting-edge model compression ideas into usable tools.
4 years of coding experience
Doctor of Philosophy - PhD, Generative (Language) Model Compression with Algebraic Approaches., Doctor of Philosophy - PhD, Generative (Language) Model Compression with Algebraic Approaches. at Imperial College London
Master of Engineering - MEng, Computer Science, Master of Engineering - MEng, Computer Science at University of Science and Technology of China
Bachelor of Engineering - BE, Computer Science, Ranked 1st in the department., Bachelor of Engineering - BE, Computer Science, Ranked 1st in the department. at China University of Petroleum, Beijing