Zhao, Mingkuan, Wentao Hu, Jiayin Wang, Xin Lai, Tianchen Huang, Yuheng Min, Rui Yan, and Xiaoyan Zhu. 2026. “Making Every Head Count: Sparse Attention Without the Speed-Performance Trade-off”. Proceedings of the AAAI Conference on Artificial Intelligence 40 (41):34959-67. https://doi.org/10.1609/aaai.v40i41.40800.