Yuan Cheng

Research Interests

My research interests lie in Learning Theory, Optimization, Reinforcement Learning Theory.

Publication (* denotes equal contribution)

• In-Context Convergence of Transformers. Y. Huang, Y. Cheng, Y. Liang. NeurIPS 2023 Mathematics of Modern Machine Learning Workshop (Oral), ICML 2024.
• Provable Benefits of Multi-task RL under Non-Markovian Decision Making Processes. R. Huang*, Y. Cheng*, J. Yang, V. Tan, Y. Liang. International Conference on Learning Representations (ICLR) 2024..
• Sample Complexity Characterization for Linear Contextual MDPs. J. Deng, Y. Cheng, S. Zou, Y. Liang. International Conference on Artificial Intelligence and Statistics (AISTATS) 2024.
• Provably Efficient Algorithm for Nonstationary Low-Rank MDPs. Y. Cheng, J. Yang, Y. Liang. Conference on Neural Information Processing Systems (NeurIPS) 2023.
• Online Min-max Problems with Non-convexity and Non-stationarity. Y. Huang*, Y. Cheng*, Y. Liang and L. Huang. Transactions on Machine Learning Research (TMLR), 2023.
• Improved Sample Complexity for Reward Free Reinforcement Learning under Low-rank MDPs. Y. Cheng*, R. Huang*, J. Yang and Y. Liang. International Conference on Learning Representations (ICLR) 2023.
• Provable Benefit of Multitask Representation Learning in Reinforcement Learning. Y. Cheng, S. Feng, J. Yang, H. Zhang, Y. Liang. Conference on Neural Information Processing Systems (NeurIPS) 2022 , Spotlight.

Preprint (* denotes equal contribution)