Yuxuan Li
Technical Partner, ModelBest; Chief Scientist, Intelligent Evolution
Yuxuan Li is a postdoctoral researcher under the Shuimu Scholars Program at Tsinghua University and holds a Ph.D. in Computer Science with a specialization in High-Performance Computing. He was admitted to Tsinghua University through the National Olympiad in Informatics (NOI) recommendation program and won five championships in seven domestic and international supercomputing competitions, including the ISC'17 and ASC'17 World Championships. He previously served as a core technical lead at the National Supercomputing Wuxi Center, home of Sunway TaihuLight, where he designed key simulation algorithms for the Jiuzhang photonic quantum computer, contributing to China’s first achievement of quantum advantage. The work was published in Science as a co-authored paper. He later served as the technical lead in building an independent domestic AI computing infrastructure from the ground up. He is currently a co-first author of the MiniCPM model series and is pioneering the concept of “AI manufacturing AI,” leveraging agents to automate the end-to-end development of large language models. He has co-authored more than 30 papers in top international conferences and journals, with over 5,000 citations on Google Scholar. He has received honors including the China Youth May Fourth Medal (Collective) and the Jiangsu “U35 Explorer Award.”
Topic
AI Building AI: From a Five-Level Evolution Framework to the Autonomous Generation of AI Infrastructure
AI building AI is a core pathway toward the next generation of artificial intelligence. We propose a five-level evolution framework that characterizes the depth of AI’s involvement in AI research and development, ranging from assisted suggestions to autonomous exploration. Through practical exploration, we have developed a unifying methodology—Forge Engineering. As the cost of AI-generated code approaches zero, the goal is no longer to maximize general-purpose reuse. Instead, each specific requirement can be addressed with a dedicated, vertically integrated implementation, allowing performance to approach its theoretical limits through case-by-case specialization. Based on this methodology, we have validated the approach through three engineering practices: An AI-forged training framework that outperforms Megatron-LM and supports production-scale pre-training. An AI-forged operator library that surpasses the best publicly available solutions across individual cases. An AI-forged inference kernel that achieves higher speed than mainstream engines such as SGLang. Outline The Five-Level Evolution Framework for AI Building AI The Core Principles of Forge Engineering Three Engineering Practices: Performance breakthroughs in a training framework, operator library, and inference kernel
