Summary
Jun Yan is a Senior Research Scientist focused on advancing agentic capabilities and robustness for Gemini at Google DeepMind, drawing on nine years of research and industry experience. With a PhD in Computer Science from USC and foundational engineering training from Tsinghua, he blends deep theoretical grounding with practical system-building across major labs including Google, Meta, Amazon, and Samsung Research. His work spans agent memory, self-evolution, code agents, and post-training interventions to push large models toward reliable, autonomous behaviors. Jun has a strong internship-to-research trajectory that highlights rapid impact in applied ML products and core language understanding problems. Based in Mountain View, he maintains an academic-facing homepage showcasing his research—reflecting a rare combination of production-focused experimentation and scholarly output. Colleagues describe him as a researcher who moves quickly from conceptual ideas to robust agentic prototypes.
9 years of coding experience
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of Southern California
High School Diploma, High School Diploma at The High School Affiliated to Renmin University of China
Bachelor of Engineering - BE, Electronic Engineering, Bachelor of Engineering - BE, Electronic Engineering at Tsinghua University