KernelMosaic: An Open Dataset and Agentic Framework for CUDA Kernel Generation

KernelMosaic is an open dataset pairing diverse CUDA tasks with verified solutions and trajectories. An agentic framework guides a strong LLM through implementation, correctness repair, and performance optimization. Supervised fine-tuning of a small open-source model on KernelMosaic raised correctness on KernelBench L1, L2, and L3 to 88%, 82%, and 16% respectively.