Weizhe Hua

My background spans model architecture, large-scale pretraining, and reasoning/RL. I earned my PhD from Cornell, then worked as a researcher at Google Brain / Google DeepMind before joining Microsoft AI.

Weizhe Hua

Experience

2025-now

MAI Superintelligence

Core contributor to MAI-Thinking-1, focusing on reasoning/RL and long-horizon agentic coding.

2021-2024

Google Brain / Google DeepMind

  • Worked on efficient LLM model architecture, including Gated Attention Unit.
  • Core contributor to Veo 1 and Veo 2, focusing on video model pretraining.