QiMeng-Tensify: Scaling Up Tensor Computation Optimization via Architecture-Aware LLM-Guided MCTS.

Shouyang Dong, Jun Bi, Yuanbo Wen, Xiyue Yu, Jianxing Xu, Guanglin Xu, Ling Li, Xuehai Zhou, Tianshi Chen, Qi Guo

Browse the full ISCA paper archive.