🧑🔬 Zhengyan Zhang — papers of 2025 — Elamur Research1 paper · all years of Zhengyan ZhangNative Sparse Attention: Hardware-Aligned and Natively Trainable Sparse AttentionarXiv:2502.11089Jingyang Yuan, Huazuo Gao, Damai Dai, et al.2025 · 142 citationsThis page in Markdown → https://md.research.elamur.ai/free/en/authors/zhengyan-zhang/year/2025/