Hatchet News

Can LLMs write better GPU kernals if you keep asking them to “Optimze it for my GPU”?