About Me
I'm an Eric & Wendy Schmidt AI in Science Postdoctoral Fellow in the AI for Science Institute at Cornell University, working with Prof. Fengqi You. Before that, I received my Ph.D. in Computer Science from the Institute for Interdisciplinary Information Sciences (IIIS) at Tsinghua University in 2025, advised by Prof. Chenye Wu and Prof. Ran Duan. I was a visiting student researcher at the Computing + Mathematical Sciences (CMS) Department, Caltech from Aug. 2023 to Mar. 2024, advised by Prof. Adam Wierman. I earned my bachelor's degree in Computer Software Engineering from Huazhong University of Science & Technology in 2020.Contact: chenbei DOT lu AT cornell.edu
Research Interests
My research lies at the intersection of reinforcement learning, stochastic optimization, and robust control, with a focus on sustainable energy and computing systems.
Theoretically, I develop reinforcement learning and control algorithms that leverage problem-specific structures to overcome the curse of dimensionality in sample efficiency and provide rigorous safety guarantees, which can enable their deployment in data-limited and safety-critical environments.
Practically, I apply these designed methods to the online operation of complex networked systems: (i) designing efficient and reliable scheduling algorithms for power grids to enhance the stability and sustainability of global energy supply, and (ii) optimizing energy–computation co-scheduling in large-scale data centers to achieve globally efficient and sustainable energy–AI integration.
News
- *Sep. 18, 2025, our work 'Reinforcement Learning with Imperfect Transition Predictions: A Bellman-Jensen Approach' has been accepted by NeurIPS 2025 as a Spotlight!
- *Sep. 9, 2025, our work 'On the Optimal Deterministic Policy Learning in Chance-Constrained Markov Decision Processes' has been accepted by IEEE Control Systems Letters.
- *May. 1, 2025, our work 'Overcoming the Curse of Dimensionality in Reinforcement Learning Through Approximate Factorization' has been accepted by ICML 2025. Thanks to the excellent collaborators and look forward to seeing you in Vancouver!
- *Apr. 21, 2025, our work 'Cost-Effective Closed-Loop Bilevel Robust Optimization for Joint Chance-Constrained Economic Dispatch' has been accepted by ACM e-Energy 2025 as a full paper. Thanks to the excellent collaborators and look forward to seeing you in Rotterdam!
- *Jul. 21, 2024, I'm excited to attend IEEE PESGM 2024 in Seattle, US, where I will present our work on sample-adaptive joint chance-constrained optimization for economic dispatch. Looking forward to seeing you all!
- *Jan. 1, 2024, our work 'Self-Improving Online Storage Control for Stable Wind Power Commitment' has been accepted by IEEE Transactions on Smart Grid. Thanks to the excellent collaborators!
- *Aug. 27, 2023, I begin an exciting six-month visit to the Computing + Mathematical Sciences (CMS) Department, Caltech, collaborating with Prof. Adam Wierman and many inspiring researchers!
Selected Publications
- “Reinforcement Learning with Imperfect Transition Predictions: A Bellman-Jensen Approach“
Chenbei Lu, Zaiwei Chen, Tongxin Li, Chenye Wu, Adam Wierman
NeurIPS 2025 (Spotlight).- “Overcoming the Curse of Dimensionality in Reinforcement Learning Through Approximate Factorization“
Chenbei Lu, Laixi Shi, Zaiwei Chen, Chenye Wu, Adam Wierman
ICML 2025. [pdf]- “Self-Improving Online Storage Control for Stable Wind Power Commitment“
Chenbei Lu, Hongyu Yi, Jiahao Zhang, Chenye Wu
IEEE Transactions on Smart Grid. [pdf]- “Sample-Adaptive Robust Economic Dispatch with Statistical Guarantees”
Chenbei Lu, Nan Gu, Wenqian Jiang, Chenye Wu
IEEE Transactions on Power Systems. [pdf] - “Overcoming the Curse of Dimensionality in Reinforcement Learning Through Approximate Factorization“