Yuxuan Cao

Hakuna Matata.

Yuxuan Cao

I am a second-year Ph.D. student in Computer Science at HKUST, supervised by Prof. Junxian He. Previously, I received my master’s degree in Computer Science and Technology from Zhejiang University, advised by Prof. Yang Yang and Prof. Jiarong Xu. My work with them and Prof. Carl Yang focused on data-centric machine learning.

I want to understand how data gives rise to new capabilities—and how increasingly capable AI systems can create the data they need to improve further. I am particularly interested in physical intelligence, where agents learn through interaction and turn their experience into useful capabilities in the real world.

News

Oct 08, 2026 🚀 We introduce Mine Odyssey, a benchmark for spatial agentic intelligence in the wild, with 180 tasks across 30 real-world Minecraft maps. Check out the paper and code!
Oct 29, 2025 🚀 We introduce Toolathlon, a benchmark for diverse, realistic, and long-horizon tool use by language agents. Check out the paper and code!
Nov 20, 2024 🎉 One paper is accepted to SIGKDD 2025: How to use Graph Data in the Wild to Help Graph Anomaly Detection?
Jun 20, 2024 🎉 One paper is accepted to ACL 2024: Cross-lingual knowledge editing in large language models
Nov 15, 2023 🎉 One paper is accepted to NeurIPS 2023:Better with Less: A Data-Centric Prespective on Pre-Training Graph Neural Networks.

Publications

2026

  1. mine-odyssey.png
    Mine Odyssey: Benchmarking Spatial Agentic Intelligence in the Wild
    Yuxuan Cao, Junlong Li, Hao Li, and Junxian He
    Preprint, 2026
    PDF Code Website

2025

  1. toolathlon-tasks.png
    The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution
    Junlong Li, Wenshuo Zhao, Jian Zhao, Weihao Zeng, Haoze Wu , Xiaochen Wang, Rui Ge, Yuxuan Cao, and 13 more authors
    In International Conference on Learning Representations (ICLR 2026), 2025
    PDF Code Website

2024

  1. use.png
    How to Use Graph Data in the Wild to Help Graph Anomaly Detection?
    Yuxuan Cao, Jiarong Xu, Chen Zhao, Jiaan Wang , Carl Yang, and 2 more authors
    Proceedings of the 31th ACM SIGKDD Conference on Knowledge Discovery and Data Mining(KDD’25), 2024
  2. cross.png
    Cross-lingual knowledge editing in large language models
    Jiaan Wang, Yunlong Liang, Zengkui Sun, Yuxuan Cao, Jiarong Xu, and 1 more author
    Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (ACL), 2024

2023

  1. when.png
    When to Pre-Train Graph Neural Networks? From Data Generation Perspective!
    Yuxuan Cao*, Jiarong Xu* , Carl Yang, Jiaan Wang, Yunchao Zhang, and 3 more authors
    Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining(KDD), 2023
  2. aptmodel.png
    Better with less: A data-active perspective on pre-training graph neural networks
    Jiarong Xu, Renhong Huang, Xin Jiang, Yuxuan Cao , Carl Yang, and 2 more authors
    Advances in Neural Information Processing Systems(NeurIPS), 2023

Honors&Awards

  • Hong Kong PhD Fellowship Scheme(2025)
  • National Scholarship(2023)
  • Outstanding Graduate of Zhejiang Province(2022)
  • Zhejiang Government Scholarship(2021)
  • First Prize Scholarship of Zhejiang University(2019, 2021)

Education

  • Hong Kong University of Science and Technology
    Ph.D. in Computer Science and Engineering
    2025.09 - Present

  • Zhejiang University
    Master in Computer Science and Technology
    2022.09 - 2025.04

  • Zhejiang University
    Bachelor in Computer Science and Technology
    2018.09 - 2022.07

Miscellaneous

In my free time, I enjoy hiking and writing. I’m an avid fan of Khalil Fong, an exceptional artist and musician. I also play in a band called 'ANS' with my friends. I am interested in cognitive science, cosmology, and metaphysics, with a particular focus on exploring the interconnections between their underlying principles.