Claire Zhang

Hello! I’m Claire. I’m obsessed with understanding the dynamics of learning algorithms, and also engineering them to run at the limits of hardware.

I am a researcher on the CoreML team at Cerebras Systems, focusing on decoding algorithms to push the world’s fastest inference engine even further. My prior work covers training dynamics, optimization, and model architecture to efficiently and optimally scale foundation model training, as well as the gradient dynamics in reinforcement learning. All while having fun programming the world’s largest silicon.

Previously at Apple, I researched continual learning and worked on augmented reality systems.

I obtained my Master’s in Electrical Engineering at Stanford University. Before then, I studied Biomedical Engineering, Electrical Engineering, and Computer Science at Washington University in St. Louis, and spent a summer in Germany doing research at RWTH - ich spreche ein bisschen Deutsch.

Outside of work, I enjoy tennis, road cycling, and over-optimizing my plan for the next mountain to explore.