Biography
Dr. Xianyuan Zhan is a research associate professor at the Institute for AI Industry Research (AIR), Tsinghua University. He received a dual Master’s degree in Computer Science and Transportation Engineering, and a PhD degree in Transportation Engineering from Purdue University. Before joining AIR, Dr. Zhan was a data scientist at JD Technology and also a researcher at Microsoft Research Asia (MSRA). Dr. Zhan previously led the research and development of AI-driven industrial system optimization products at JD Technology. He has published more than 70 papers in key journals and conferences in the field of Transportation Engineering and Computer Science. He is also a reviewer for many top transportation and computer science journals and conferences. He is currently a committee member of China Computer Federation-Artificial Intelligence & Pattern Recognition (CCF-AI) Committee.
- Group Website: https://air-dream.netlify.app/
- Group Code Repository: https://github.com/AIR-DI
Research Interests
- Offline deep reinforcement learning
- Offline imitation learning
- Foundation models for decision-making
- Complex system optimization
- Autonomous driving
We are hiring!!!
Our team is looking for student interns/postdocs at AIR! If you are interested in the research directions of offline RL/IL, AI alignment/AI safety, embodied AI, decision-making in autonomous driving, please feel free to send me an e-mail at zhanxianyuan@air.tsinghua.edu.cn!
Recent News and Activities
- Jan. 2025: Our two recent papers “Robo-MUTUAL: Robotic Multimodal Task Specification via Unimodal Learning” and “H2O+: An Improved Framework for Hybrid Offline-and-Online RL with Dynamics Gaps” have been accepted in ICRA 2025!
- Jan. 2025: We have released “Universal Actions for Enhanced Embodied Foundation Models”, which learns universal actions to power any robotic embodiment, physical meaning, and control interfaces! Project page available at https://2toinf.github.io/UniAct/.
- Jan. 2025: Our three recent papers “Data Center Cooling System Optimization Using Offline Reinforcement Learning”, “Diffusion-Based Planning for Autonomous Driving with Flexible Guidance”, and “Skill Expansion and Composition in Parameter Space” have been accepted in ICLR 2025!
- Dec. 2024: Our recent paper “Are Expressive Models Truly Necessary for Offline RL?” has been accepted as an oral paper in AAAI 2025!
- Sep. 2024: Our two recent papers “Instruction-Guided Visual Masking” and “Diffusion-DICE: In-Sample Diffusion Guidance for Offline Reinforcement Learning” have been accepted in NeurIPS 2024!
- Jul. 2024: Our two recent papers “DecisionNCE: Embodied Multimodal Representations via Implicit Preference Learning” and “Instruction-Guided Visual Masking” have won the Outstanding Paper Awards at ICML 2024 Workshop on Multi-modal Foundation Model meets Embodied AI (MFM-EAI).
- May. 2024: Our four recent papers “DecisionNCE: Embodied Multimodal Representations via Implicit Preference Learning”, “OMPO: A Unified Framework for Reinforcement Learning under Policy and Dynamics Shifts”, “Offline-Boosted Actor-Critic: Adaptively Blending Optimal Historical Behaviors in Deep Off-Policy RL”, “Seizing Serendipity: Exploiting the Value of Past Success in Off-Policy Actor-Critic” have been accepted in ICML 2024!
- Apr. 2024: Our recent survey paper “A Comprehensive Survey of Cross-Domain Policy Transfer for Embodied Agents” has been accepted in IJCAI 2024.
- Jan. 2024: Our four recent papers “Revealing the Mystery of Distribution Correction Estimation via Orthogonal-gradient Update”, “Safe Offline Reinforcement Learning with Feasibility-Guided Diffusion Model”, “Query-Policy Misalignment in Preference-Based Reinforcement Learning”, and “OpenChat: Advancing Open-source Language Models with Mixed-Quality Data” have been accepted in ICLR 2024!