Publications

Selected publications and recent work in reinforcement learning, embodied AI, and vision-language-action models.

Selected Publications

Representative papers that are most aligned with my current research direction.

Full List

A broader list of papers spanning reinforcement learning, embodied manipulation, and VLA-related research.

2026

ALOE: Action-Level Off-Policy Evaluation for Vision-Language-Action Model Post-Training
Selected
R. Yang, H. Wang, C. Liu, X. Yan, Y. Wang, X. Du, S. Yue, Y. Liu, C. Zhang, et al.
arXiv preprint · Demo
VINE: Taming Generative Control Policies for Reinforcement Learning
Selected
R. Yang, Z. Han, H. Li, H. Wang, Z. Wu, R. Zhang, Z. Zhang, Z. Chen, X. Yan, C. Liu, Y. Chen, W. Shan, M. Yao
arXiv preprint · Demo

2025

How Do VLAs Effectively Inherit from VLMs?
C. Zhang, R. Yang, X. Chen, K. Wang, L. Zhao, Y. Chen, J. Bian
arXiv preprint
Supervised Optimism Correction: Be Confident When LLMs Are Sure
J. Zhang, R. Yang, S. Liu, T. E. Lin, F. Huang, Y. Chen, Y. Li, D. Tao
Association for Computational Linguistics (ACL)
Temporal Consistent Multi-View Perception for Robust Embodied Manipulation
H. Chen, R. Yang, J. Zhang, X. Wen, Y. Chen, D. Yu, C. Bai, Z. Wang
Pattern Recognition
Beyond Human Demonstrations: Diffusion-Based Reinforcement Learning to Generate Data for VLA Training
Selected
R. Yang, H. Wei, R. Zhang, Z. Feng, X. Chen, T. Li, C. Zhang, L. Zhao, J. Bian, et al.
arXiv preprint
Discover, Learn, and Reinforce: Scaling Vision-Language-Action Pretraining with Diverse RL-Generated Trajectories
Selected
R. Yang, Z. Feng, T. Zhang, K. Wang, C. Zhang, L. Zhao, X. Su, Y. Chen, et al.
arXiv preprint

2024

Constrained Ensemble Exploration for Unsupervised Skill Discovery
Selected
C. Bai, R. Yang, Q. Zhang, K. Xu, Y. Chen, T. Xiao, X. Li
International Conference on Machine Learning (ICML)

2023

Behavior Contrastive Learning for Unsupervised Skill Discovery
Selected
R. Yang, C. Bai, H. Guo, S. Li, B. Zhao, Z. Wang, P. Liu, X. Li
International Conference on Machine Learning (ICML)

For the most up-to-date list, please also see my Google Scholar profile.