加州理工学院 Bren 讲席教授 · 前 NVIDIA AI 研究高级总监Bren Professor at Caltech · ex-NVIDIA Sr. Director of AI Research
deep-learningsystems
加州理工学院计算科学 Bren 讲席教授，AI for Science 方向的代表人物：其团队提出的神经算子（含傅里叶神经算子 FNO）学习函数空间之间的映射，让基于偏微分方程的仿真提速数个数量级，也是 FourCastNet——首个可与数值预报比肩的深度学习全球天气模型——的基础。2018–2023 年任 NVIDIA AI 研究高级总监，此前是 AWS 首席科学家。Accelerated Understanding 联合创始人，将这套方法用于科学发现。
Bren Professor of Computing at Caltech and one of the leading figures in AI for physical sciences: her group introduced neural operators (including the Fourier Neural Operator) that learn mappings between function spaces, enabling orders-of-magnitude speedups in PDE-based simulation — the basis of FourCastNet, the first deep-learning global weather model to rival numerical forecasting. Senior Director of AI Research at NVIDIA from 2018 to 2023, earlier a principal scientist at AWS. Co-founder of Accelerated Understanding, applying these methods to scientific discovery.