About

I am a second-year Ph.D. student at the School of Information and Control Engineering, China University of Mining and Technology (CUMT), born on March 2, 2001. My research focuses on offline reinforcement learning, goal-conditioned reinforcement learning, and generative models. I am currently transitioning to Vision-Language-Action (VLA) models to explore the mysteries of embodied intelligence. I am advised by Prof. Xuesong Wang.

Research Interests

  • Offline Reinforcement Learning
  • Goal-Conditioned Reinforcement Learning
  • Generative Models

News

  • [Nov 2026]One paper released on arXiv as a preprint.
  • [Sep 2026]One paper accepted at NeurIPS 2026 as Poster.
  • [Aug 2026]One paper accepted at IEEE/CAA Journal of Automatica Sinica.
  • [May 2026]One paper accepted at ICML 2026 as Poster.
  • [Jul 2025]One paper accepted at IEEE Transactions on Systems, Man, and Cybernetics: Systems.
  • [Apr 2025]One paper accepted at Neural Networks.
  • [Feb 2025]One paper accepted at IEEE Transactions on Cognitive and Developmental Systems.
  • [Aug 2024]One paper accepted at Control and Decision.

Publications

Marginal-Preserving Coupled Diffusion Composition for Long-Horizon Planning

Hengrui Zhang, Qiyu Wang, Yuhu Cheng, Brooks Paige, Xuesong Wang

arXiv preprint arXiv:2611.15585, 2026

Diffusion Subgoal Planning for Long-Horizon Offline Goal-Conditioned Reinforcement Learning

Hengrui Zhang, Yuhu Cheng, C. L. Philip Chen, Xuesong Wang

Conference on Neural Information Processing Systems (NeurIPS), 2026 (Poster)

Return-Critic: Bridging Goal Discrepancy for Efficient Visual Reinforcement Learning

Ruyi Lu, Xuesong Wang, Hengrui Zhang, Yuhu Cheng

International Conference on Machine Learning (ICML), 2026 (Poster)

Flow Matching Regularization for Offline Reinforcement Learning

Jiazhi Zhang, Yuhu Cheng, Hengrui Zhang, Xuesong Wang

IEEE/CAA Journal of Automatica Sinica, accepted 2026

PCDT: Pessimistic Critic Decision Transformer for Offline Reinforcement Learning

Xuesong Wang, Hengrui Zhang, Jiazhi Zhang, C. L. Philip Chen, Yuhu Cheng

IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2025, 55(10): 7247–7258

Visual Reinforcement Learning Based on Multiview Optimization Aggregation

Xuesong Wang, Ruyi Lu, Hengrui Zhang, Yuhu Cheng

IEEE Transactions on Cognitive and Developmental Systems, 2025, 17(4): 1011–1021

Diffusion Policy Distillation for Offline Reinforcement Learning

Jiazhi Zhang, Yuhu Cheng, C. L. Philip Chen, Hengrui Zhang, Xuesong Wang

Neural Networks, 2025, 190: 107694

Offline Reinforcement Learning Based on Advantage-Constrained Diffusion Policy

Xuesong Wang, Hengrui Zhang, Jiazhi Zhang, Yuhu Cheng

Control and Decision, 2025, 40(06): 1903–1912

Education

Ph.D. in Control Science and Engineering
China University of Mining and Technology
2025 — Present
M.Sc. in Control Science and Engineering
China University of Mining and Technology
2023 — 2025
B.Sc. in Internet of Things Engineering
Yancheng Institute of Technology
2019 — 2023

Honors & Awards

Ph.D.

  • First-Class Entrance Scholarship

M.Sc.

  • Second-Class Entrance Scholarship
  • Second-Class Academic Scholarship

B.Sc.

  • National Scholarship
  • National Endeavor Scholarship
  • Yancheng "Yellow Sea Pearl" Scholarship
  • First-Class Academic Scholarship ×8
  • National Second Prize, 17th "Challenge Cup"
  • National Third Prize, China College Student Computer Design Competition
  • National Second Prize, China College Student Computer Design Competition (Preliminary)
@article{zhang2026mcdc,
  title={Marginal-Preserving Coupled Diffusion Composition for Long-Horizon Planning},
  author={Zhang, Hengrui and Wang, Qiyu and Cheng, Yuhu and Paige, Brooks and Wang, Xuesong},
  journal={arXiv preprint arXiv:2611.15585},
  year={2026}
}
@inproceedings{zhang2026diffusion,
  title={Diffusion Subgoal Planning for Long-Horizon Offline Goal-Conditioned Reinforcement Learning},
  author={Zhang, Hengrui and Cheng, Yuhu and Chen, C. L. Philip and Wang, Xuesong},
  booktitle={Advances in Neural Information Processing Systems},
  year={2026}
}
@inproceedings{lu2026returncritic,
  title={Return-Critic: Bridging Goal Discrepancy for Efficient Visual Reinforcement Learning},
  author={Lu, Ruyi and Wang, Xuesong and Zhang, Hengrui and Cheng, Yuhu},
  booktitle={International Conference on Machine Learning},
  year={2026}
}
@article{zhang2026flow,
  title={Flow Matching Regularization for Offline Reinforcement Learning},
  author={Zhang, Jiazhi and Cheng, Yuhu and Zhang, Hengrui and Wang, Xuesong},
  journal={IEEE/CAA Journal of Automatica Sinica},
  year={2026}
}
@article{wang2025pcdt,
  title={PCDT: Pessimistic Critic Decision Transformer for Offline Reinforcement Learning},
  author={Wang, Xuesong and Zhang, Hengrui and Zhang, Jiazhi and Chen, C. L. Philip and Cheng, Yuhu},
  journal={IEEE Transactions on Systems, Man, and Cybernetics: Systems},
  volume={55},
  number={10},
  pages={7247--7258},
  year={2025},
  doi={10.1109/TSMC.2025.3583392}
}
@article{wang2025multiview,
  title={Visual Reinforcement Learning Based on Multiview Optimization Aggregation},
  author={Wang, Xuesong and Lu, Ruyi and Zhang, Hengrui and Cheng, Yuhu},
  journal={IEEE Transactions on Cognitive and Developmental Systems},
  volume={17},
  number={4},
  pages={1011--1021},
  year={2025},
  doi={10.1109/TCDS.2025.3540115}
}
@article{zhang2025diffusion,
  title={Diffusion Policy Distillation for Offline Reinforcement Learning},
  author={Zhang, Jiazhi and Cheng, Yuhu and Chen, C. L. Philip and Zhang, Hengrui and Wang, Xuesong},
  journal={Neural Networks},
  volume={190},
  pages={107694},
  year={2025},
  doi={10.1016/j.neunet.2025.107694}
}
@article{wang2025advantage,
  title={Offline Reinforcement Learning Based on Advantage-Constrained Diffusion Policy},
  author={Wang, Xuesong and Zhang, Hengrui and Zhang, Jiazhi and Cheng, Yuhu},
  journal={Control and Decision},
  volume={40},
  number={6},
  pages={1903--1912},
  year={2025},
  doi={10.13195/j.kzyjc.2024.0618}
}