ZHU LEI (祝磊)

I’m a researcher at Huawei Noah’s Ark Lab (Hong Kong). I got my PhD degree from CityUHK under the supervision of Prof. Rynson W.H. Lau in July, 2024. Before that, I received my bachelor degree in Computer Science and Engineering from Dalian University of Technology (rank 1/102) and spent one year as PhD student at CUHK under the supervision of Eric Lo.

My research focus on long sequence modelling, efficient deep learning, and agentic AI.

News

Sep 19, 2025	One paper on MoE speculative decoding has been accepted by NeurIPS 2025 as spotlight.
Sep 12, 2025	We have open sourced our multi-agent deep research project DeepDiverV2.
Aug 21, 2025	One paper on latency-aware test time scaling is accepted by EMNLP Findings.
Nov 14, 2024	Invited talk by UCSD Hao AI Lab about “Efficient Attention Mechanisms”.
Oct 30, 2024	I’m pleased to receive NeurIPS travel award and be selected as Top Reviewers.

Selected Publications

ACL

RelayAttention for Efficient Large Language Model Serving with Long System Prompts

Lei Zhu, Xinjiang Wang, Wayne Zhang, and 1 more author

In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , 2024

arXiv Bib Code

@inproceedings{zhu2023relayattention,
  title = {RelayAttention for Efficient Large Language Model Serving with Long System Prompts},
  author = {Zhu, Lei and Wang, Xinjiang and Zhang, Wayne and Lau, Rynson WH},
  booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2024},
}

CVPR

BiFormer: Vision Transformer with Bi-Level Routing Attention

Lei Zhu, Xinjiang Wang, Zhanghan Ke, and 2 more authors

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023

arXiv Bib Code Poster

@inproceedings{zhu2023biformer,
  title = {BiFormer: Vision Transformer with Bi-Level Routing Attention},
  author = {Zhu, Lei and Wang, Xinjiang and Ke, Zhanghan and Zhang, Wayne and Lau, Rynson WH},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages = {10323--10333},
  year = {2023},
  google_scholar_id = {d1gkVwhDpl0C},
}