KernelBench: Can LLMs Write Efficient GPU Kernels?

Feb 2025