Darren Wu
I currently work on midtraining (RL) scaling and inference performance at Google DeepMind. I was a core contributor to Gemini 2.5, 3.1, and 4+.
Before that, I interned at Stripe and Google Brain.
I studied Electrical Engineering, Computer Science, and Mathematics at Duke University. There, I was part of Duke Electric Vehicles and researched the influence of air pollution on climate and human health under Dr. Mike Bergin.
I've been an open-source contributor to JAX, Pallas, XLA, PyTorch, TensorFlow, vLLM, LLVM, and more.