I am Current a Second-year Master student in ShanghaiTech University under the supervision of Prof. Ye Shi and Prof. Jingya Wang

My Research interest include Humanoid Locomotion and Humanoid Loco-manipulation. Now I focus on the topic of Simulated synthetic data and Humanoid robot system. Feel free to contact me if you are interested in these fantasic topics!

Email : guitx2025 AT shanghaitech DOT edu DOT cn

📝 Publications

Google Scholar

# Equal contribution · * Corresponding author

2026

arXiv 2026
SUGARv2 interaction reference synthesis and visual humanoid policy learning framework

SUGARv2: Learning Diverse Visual Humanoid Loco-Manipulation Skills from Synthesized Interaction References

Tianshu Wu, Tianxiang Gui, Yue Chen, Tingxu Liu, Ziye Peng, Guangjie Yang, Xiangqi Kong, Ye Shi, Jingya Wang, Hao Dong

arXiv preprint, 2026.

[paper] [project]

arXiv 2026
InstAgent overview for heterogeneous robot collaboration

InstAgent: A Recursively Self-Improving Agent Operating System for Heterogeneous Robot Collaboration

Haoyang Luo#, Junhao Qin#, Jingyan Cao#, Yi Huang, Guangjie Yang, Xiaoyu Chen, Haotao Lu, Tianxiang Gui, Ye Shi, Jingya Wang*

arXiv preprint, 2026.

[paper] [project]

arXiv 2026
SteerGenPO latent steering framework for stable robotic control

Steering Generative Reinforcement Learning into Stable Robotic Controller

Yixuan Wang#, Shutong Ding#, Ke Hu#, Tianxiang Gui, Jingya Wang, Ye Shi*

arXiv preprint, 2026.

[paper]

2025

Sisyphus: An attention-driven humanoid object interaction policy

Tianxiang Gui, Ke Yang, Yahao Fan, Chixuan Zhang, Jingya Wang*, Ye Shi*

IROS 2025 Workshop: Augmentative Human–Robot Interaction.

arXiv 2025
DreamPolicy humanoid locomotion across diverse terrains

DreamPolicy: A Unified World-model Policy for Scalable Humanoid Locomotion

Yahao Fan#, Tianxiang Gui#, Kaiyang Ji#, Shutong Ding, Chixuan Zhang, Yifeng Xu, Ke Yang, Jiayuan Gu, Jingyi Yu, Jingya Wang*, Ye Shi*

arXiv preprint, 2025 (updated 2026).

Originally titled One Policy but Many Worlds: A Scalable Unified Policy for Versatile Humanoid Locomotion.

[paper] [original version] [project]

Project

[SuperMimic-Retargeted] An unofficial adaptation of SuperMimic-Reproduced for humanoid motion tracking on G1, A3, and X2 Ultra in simulation, with motion retargeting, training, evaluation, and visualization workflows.

[Humanoid Terrain Bench] Multi challenging terrain for humanoid robot.

📖 Educations

  • 2025.09 - now, ShanghaiTech University, School of Information Science and Technology, major in Computer Technology
  • 2021.09 - 2025.06, SouthWest JiaoTong University, major in Electronic information

💻 Internship

  • 2025.1-2025.9, Visiting Student, ShanghaiTech University, VDI Center, YesAI Lab

🏆 Awards