Large language models have recently shown strong potential for automatically optimizing GPU kernels, yet their effectiveness depends critically on the hardwa…
机构:人大
来源:arXiv 2610.36576 | AI4Papers 论文推荐平台