avatar

Yong Liu

Ph.D. Student
Tsinghua University
liuyong1095556447@gmail.com


About Me

I am currently an AI researcher @ ByteDance. Previously, I received my Ph.D. from the School of Software at Tsinghua University, advised by Prof. Mingsheng Long at THUML.

My research interests cover Deep Learning and Foundation Models. I am currently working on generic time series foundation models and agentic forecasting. My research aims to contribute to the advancement of intelligent systems capable of handling massive and complicated temporal data across domains.

For more information, please take a look at my Google Scholar and GitHub.

News

Selected Publications

  1. arxiv
    Timer-S1: A Billion-Scale Time Series Foundation Model with Serial Scaling
    Yong Liu*, Xingjian Su*, Shiyu Wang*, Haoran Zhang*, Haixuan Liu, Yuxuan Wang, Zhou Ye, Yang Xiang, Jianmin Wang, Mingsheng Long#
    arxiv preprint
    arXiv Code Technical Report

  2. ICML
    Sundial: A Family of Highly Capable Time Series Foundation Models
    Yong Liu*, Guo Qin*, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long#
    International Conference on Machine Learning, 2025.
    arXiv Code Poster Slides Oral Paper

  3. ICML
    Timer: Generative Pre-trained Transformers Are Large Time Series Models
    Yong Liu*, Haoran Zhang*, Chenyu Li*, Xiangdong Huang, Jianmin Wang, Mingsheng Long#
    International Conference on Machine Learning, 2024.
    arXiv Code Poster Slides Dataset Accepted

  4. ICLR
    iTransformer: Inverted Transformers Are Effective for Time Series Forecasting
    Yong Liu*, Tengge Hu*, Haoran Zhang*, Haixu Wu, Shiyu Wang, Lintao Ma, Mingsheng Long#
    International Conference on Learning Representations, 2024.
    arXiv Code Poster Slides Spotlight Paper, PaperDigest Most Influential Paper

  5. NeurIPS
    Non-stationary Transformers: Exploring the Stationarity in Time Series Forecasting
    Yong Liu*, Haiwu Wu*, Jianmin Wang, Mingsheng Long#
    Conference on Neural Information Processing Systems, 2022.
    arXiv Code Poster Slides Accepted

  6. ICLR
    Timer-XL: Long-Context Transformers for Unified Time Series Forecasting
    Yong Liu*, Guo Qin*, Xiangdong Huang, Jianmin Wang, Mingsheng Long#
    International Conference on Learning Representations, 2025.
    arXiv Code Poster Slides Accepted

  7. NeurIPS
    Koopa: Learning Non-stationary Time Series Dynamics with Koopman Predictors
    Yong Liu*, Chenyu Li*, Jianmin Wang, Mingsheng Long#
    Conference on Neural Information Processing Systems, 2023.
    arXiv Code Poster Slides Accepted

  8. ICLR
    TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis
    Haiwu Wu*, Tengge Hu*, Yong Liu*, Hang Zhou, Jianmin Wang, Mingsheng Long#
    International Conference on Learning Representations, 2023.
    arXiv Code Slides Accepted, PaperDigest Most Influential Paper

  9. NeurIPS
    AutoTimes: Autoregressive Time Series Forecasters via Large Language Models
    Yong Liu*, Guo Qin*, Xiangdong Huang, Jianmin Wang, Mingsheng Long#
    Conference on Neural Information Processing Systems, 2024.
    arXiv Code Poster Slides Accepted

  10. arxiv
    Deep Time Series Models: A Comprehensive Survey and Benchmark
    Yuxuan Wang*, Haixu Wu*, Jiaxiang Dong, Yong Liu, Mingsheng Long, Jianmin Wang#
    arxiv preprint
    arXiv arXiv preprint, 2025

  11. JMLR
    Ranking and Tuning Pre-trained Models: A New Paradigm for Exploiting Model Hubs
    Kaichao You*, Yong Liu*, Ziyang Zhang, Jianmin Wang, Michael I. Jordan, Mingsheng Long#
    Journal of Machine Learning Research, 2022.
    arXiv Code Accepted

(* Equal Contribution, # Corresponding Author)

Selected Projects

Hugging Face Models

Deep Models for Time Series

Algorithms and Model Packages

Systems and Applications

Invited Talks

Services

Academic Services

Teaching Experience

Experience & Education

Selected Awards


Powered by Jekyll and Minimal Light theme.