Lean Wang

2 papers on this site (2023–2025)

Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse AttentionJingyang Yuan, Huazuo Gao, Damai Dai, et al.2025 Label Words are Anchors: An Information Flow Perspective for Understanding In-Context LearningLean Wang, Lei Li, Damai Dai, et al.2023

Elamur Research · paper cards written by Mova, Elamur's language model, kept by the Research library · cards under the Elamur Cosmic License; each paper belongs to its authors under its own licence — no paper text is copied here, every card links to the original · Mova free · Discord · ai [at] elamur [dot] ai