I am an LLM researcher at BandAI, Bytedance. I received my PhD and MPhil in Computer Science from the University of Sydney in 2026 and 2022, advised by Prof. Dacheng Tao. Prior to that, I received a BEng from the School of Automation Science and Electrical Engineering, Beihang University in 2019.

My research broadly focuses on LLM alignment, post-training, and reinforcement learning, with the goal of pushing the capabilities of frontier language models.

Token Usage

Publications

* Co-first authors

Published

  • NeurIPS

    Stabilizing Off-policy LLM Optimization with Prefix Importance Ratio

    Shiye Lei, Zhihao Cheng, and Dacheng Tao

  • ICML

    EAPO: Enhancing Policy Optimization with On-Demand Expert Assistance

    Siyao Song, Cong Ma, Zhihao Cheng, Shiye Lei, Minghao Li, Ying Zeng, Huaixiao Tou, and Kai Jia

  • KDD

    Offline Behavioral Data Selection

    Shiye Lei, Zhihao Cheng, and Dacheng Tao

  • IJCV

    Image Captions are Natural Prompts for Training Data Synthesis

    Shiye Lei*, Hao Chen*, Sen Zhang, Bo Zhao, and Dacheng Tao

  • TNNLS

    Attentive Learning Facilitates Generalization of Neural Networks

    Shiye Lei, Fengxiang He, Haowen Chen, and Dacheng Tao

  • TNNLS

    Understanding Deep Learning via Decision Boundary

    Shiye Lei, Fengxiang He, Yancheng Yuan, and Dacheng Tao

  • NeurIPS

    Offline Behavior Distillation

    Shiye Lei, Sen Zhang, and Dacheng Tao

  • TPAMI

    A Comprehensive Survey of Dataset Distillation

    Shiye Lei and Dacheng Tao

  • AIJ

    Spectral complexity-scaled generalisation bound of complex-valued neural networks

    Haowen Chen, Fengxiang He, Shiye Lei, and Dacheng Tao

Preprints

  • Preprint

    Revisiting LLM Reasoning via Information Bottleneck

    Shiye Lei, Zhihao Cheng, Kai Jia, and Dacheng Tao

  • Preprint

    EarthSynth: Generating Informative Earth Observation with Diffusion Models

    Jiancheng Pan*, Shiye Lei*, Yuqian Fu, Jiahao Li, Yanxing Liu, Yuze Sun, Xiao He, Long Peng, Xiaomeng Huang, and Bo Zhao

  • Preprint

    State Diversity Matters in Offline Behavior Distillation

    Shiye Lei, Zhihao Cheng, and Dacheng Tao

  • Preprint

    Spatial-Temporal-Fusion BNN: Variational Bayesian Feature Layer

    Shiye Lei, Zhuozhuo Tu, Leszek Rutkowski, Feng Zhou, Li Shen, Fengxiang He, and Dacheng Tao

  • Preprint

    Neural networks behave as hash encoders: An empirical study

    Fengxiang He*, Shiye Lei*, Jianmin Ji, and Dacheng Tao

Services

Conference Reviewer

ICML, NeurIPS, ICLR, AISTATS, CVPR, ICCV, AAAI, ACM MM, etc.

Journal Reviewer

TPAMI, IJCV, JMLR, etc.