About
I’m a third-year Ph.D. candidate at The Chinese University of Hong Kong, Shenzhen (CUHK-SZ),
supervised by Prof. Guiliang Liu
and co-supervised by Prof. Hongyuan Zha.
Previously, I received my Bachelor's degree (rank: 1/225) with the highest honors from
Northwestern Polytechnical University (NWPU) in 2023.
My work has been published in top-tier international AI venues such as ICLR, ICML, and NeurIPS. I am also one of the core developers to EmbodiChain
, an end-to-end,
GPU-accelerated, and modular platform for building generalized embodied intelligence.
My research focuses on:
My work has been published in top-tier international AI venues such as ICLR, ICML, and NeurIPS. I am also one of the core developers to EmbodiChain
, an end-to-end,
GPU-accelerated, and modular platform for building generalized embodied intelligence.
My research focuses on:
- Risk-sensitive / Robust Reinforcement Learning
- Reinforcement Learning Applications (e.g., Embodied AI, Sports Science)
News
- [2026.04] TacticGen was released on arXiv.
- [2026.01] One paper was accepted to MIT SSAC 2026.
- [2026.01] One paper was accepted to ICLR 2026.
- [2025.09] One paper was accepted to NeurIPS 2025.
- [2025.07] One book chapter was published online by Springer Nature in both English and German.
- [2025.01] Two papers were accepted to ICLR 2025.
- [2025.01] One paper was accepted to TMLR.
- [2024.05] One paper was accepted to ICML 2024.
- [2024.01] Two papers were accepted to ICLR 2024.
Publications
* indicates equal contribution, # indicates project leader
TacticGen: Grounding Adaptable and Scalable Generation of Football Tactics
Risk, Reward, and Reinforcement Learning in Ice Hockey Analytics
- TacticGen: Grounding Adaptable and Scalable Generation of Football Tactics
Sheng Xu, Guiliang Liu, Tarak Kharrat, Yudong Luo, Mohamed Aloulou, Javier López Peña, Konstantin Sofeikov, Adam Reid, Paul Roberts, Steven Spencer, Joe Carnall, Ian McHale, Oliver Schulte, Hongyuan Zha, Wei-Shi Zheng
arXiv preprint, 2026
[Paper] [Code] [Project Page] - Sim2Real VLA: Zero-Shot Generalization of Synthesized Skills to Realistic Manipulation
Runyi Zhao, Sheng Xu, Ruixing Jin, Yueci Deng, Yunxin Tai, Kui Jia, Guiliang Liu
International Conference on Learning Representations (ICLR), 2026
[Paper] [Code] - HoopEval: Individual Player Action Evaluation via Deep Reinforcement Learning
Xing Wang, Yu Fu, Sheng Xu, Konstantinos Pelechrinis, Mingxin Zhang, Miguel Ángel Gómez Ruano, Guiliang Liu, Shaoliang Zhang
MIT Sloan Sports Analytics Conference (SSAC), 2026
[Paper]
- A Distributional Approach to Uncertainty-Aware Preference Alignment Using Offline Demonstrations
Sheng Xu, Bo Yue, Hongyuan Zha, Guiliang Liu
International Conference on Learning Representations (ICLR), 2025
[Paper] [Code] - Risk, Reward, and Reinforcement Learning in Ice Hockey Analytics
Sheng Xu, Oliver Schulte, Yudong Luo, Pascal Poupart, Guiliang Liu
Book Chapter in Artificial Intelligence and Machine Learning in Sports Science, Springer, 2025
German Version: Künstliche Intelligenz und maschinelles Lernen in der Sportwissenschaft
[Paper (English)] [Paper (German)] [Code] - Uncertainty-aware Preference Alignment for Diffusion Policies
Runqing Miao*, Sheng Xu*#, Runyi Zhao, Wai Kin Victor Chan, Guiliang Liu
Annual Conference on Neural Information Processing Systems (NeurIPS), 2025
[Paper] [Code] - Toward Exploratory Inverse Constraint Inference with Generative Diffusion Verifiers
Runyi Zhao*, Sheng Xu*, Bo Yue, Guiliang Liu
International Conference on Learning Representations (ICLR), 2025
[Paper] [Code] - A Comprehensive Survey on Inverse Constrained Reinforcement Learning
Guiliang Liu, Sheng Xu, Shicheng Liu, Ashish Gaurav, Sriram Ganapathi Subramanian, Pascal Poupart
Transactions on Machine Learning Research (TMLR), 2025
[Paper] [Code]
- Robust Inverse Constrained Reinforcement Learning under Model Misspecification
Sheng Xu, Guiliang Liu
International Conference on Machine Learning (ICML), 2024
[Paper] [Code] - Uncertainty-aware Constraint Inference in Inverse Constrained Reinforcement Learning
Sheng Xu, Guiliang Liu
International Conference on Learning Representations (ICLR), 2024
[Paper] [Code] - Enhancing Human-AI Collaboration Through Logic-Guided Reasoning
Chengzhi Cao, Yinghao Fu, Sheng Xu, Ruimao Zhang, Shuang Li
International Conference on Learning Representations (ICLR), 2024
[Paper]
Education
Sep. 2019 – Jul. 2023
B.S. in Computer Science and Technology
(with the highest honors)
Honors and Awards
- National Scholarship 2021 & 2022
- Ph.D. Presidential Fellowship (highest scholarship), 2023-2028
- Duan Yong Ping Meritorious Research Award, 2024
- Duan Yong Ping Travel Award, 2024
- Outstanding Graduate of Shaanxi Province, 2023
- NWPU Top Undergraduate Student Award (highest honor of undergraduates, only 10 winners), 2022
- NWPU Top Class Scholarship (highest scholarship, only 10 winners) 2022
- Modern Scientists Scholarship (top 1%), 2022
- Huawei Scholarship (top 1%), 2022
Internship
Sep. 2025 – Present
Embodied AI Research Intern,
DexForce Technology, Shenzhen
Core contributor to EmbodiChain, an open-source, GPU-accelerated embodied intelligence platform; built simulation-based task pipelines and end-to-end Sim-to-Real VLA systems for robust real-robot deployment.
Sep. 2024 – Aug. 2025
Football AI Research Intern,
Birmingham City FC, UK (online)
Conducted research with
Real Analytics
on offline reinforcement learning for football analytics and recommendation.
Teaching
- DDA4230 Reinforcement Learning: Fall 2023 & 2024
Leading Teaching Assistant x 2 - DDA2001 Introduction to Data Science: Spring 2024 & 2025
Leading Teaching Assistant x 1 (Recipient of the Excellent TA Award), Teaching Assistant x 1
Service
- Conference Reviewer: NeurIPS (Top Reviewer), ICLR, ICML, ECCV, AAAI, AISTATS
- Journal Reviewer: Transactions on Machine Learning Research, Expert Systems With Applications, Pattern Recognition, Applied Intelligence
- Volunteer: ICML 2024