Publications

BILEVEL DECENTRALIZED MULTI-AGENT LEARNING
Siliang Zeng, Songtao Lu, Xiaodong Cui, Mark S Squillante, Lior Horesh, Brian ED Kingsbury, Mingyi Hong
US Patent Application No. 20250005324

From Demonstrations to Rewards: Alignment Without Explicit Human Preferences
Siliang Zeng, Yao Liu, Huzefa Rangwala, George Karypis, Mingyi Hong, Rasool Fakoor

Bridging the Training-Inference Gap in LLMs by Leveraging Self-Generated Tokens
Zhepeng Cen, Yao Liu, Siliang Zeng, Pratik Chaudhari, Huzefa Rangwala, George Karypis, Rasool Fakoor
Transactions on Machine Learning Research

Structural Estimation of Markov Decision Processes in High-Dimensional State Space with Finite-Time Guarantees
Siliang Zeng, Mingyi Hong, Alfredo Garcia
Operations Research.

On the Divergence of Decentralized Non-Convex Optimization
Mingyi Hong, Siliang Zeng, Junyu Zhang, Haoran Sun
SIAM Journal on Optimization

Reinforcement Learning in Inference Time: A Perspective from Successive Policy Iterations.
Xinnan Zhang, Chenliang Li, Siliang Zeng, Jiaxiang Li, Zhongruo Wang, Songtao Lu, Alfredo Garcia, Mingyi Hong
Reasoning and Planning for LLMs workshop @ ICLR2025

Joint Demonstration and Preference Learning Improves Policy Alignment with Human Feedback
Chenliang Li, Siliang Zeng, Zeyi Liao, Jiaxiang Li, Dongyeop Kang, Alfredo Garcia, Mingyi Hong
The Thirteenth International Conference on Learning Representations (ICLR 2025). (Spotlight: 5.1% )
Also accepted to NeurIPS 2025 Workshop on Fine-Tuning in Modern Machine Learning: Principles and Scalability (FITML 2025).

Understanding Inverse Reinforcement Learning under Overparameterization: Non-Asymptotic Analysis and Global Optimality
Ruijia Zhang, Siliang Zeng, Chenliang Li, Alfredo Garcia, Mingyi Hong
The 28th International Conference on Artificial Intelligence and Statistics (AISTATS 2025)

Policy Optimization can be Memory-Efficient: LLM Alignment Through Successive Policy Re-weighting (SPR).
Xinnan Zhang*, Siliang Zeng*, Jiaxiang Li*, Kaixiang Lin, Mingyi Hong
NeurIPS 2024 Workshop on Fine-Tuning in Modern Machine Learning: Principles and Scalability (FITML)

Getting More Juice Out of the SFT Data: Reward Learning from Human Demonstration Improves SFT for LLM Alignment
Jiaxiang Li, Siliang Zeng, Hoi-To Wai, Chenliang Li, Alfredo Garcia, Mingyi Hong
Thirty-eighth Annual Conference on Neural Information Processing Systems (NeurIPS 2024).
Also accepted to ICML 2024 Workshop on Theoretical Foundations of Foundation Models.

When Demonstrations Meet Generative World Models: A Maximum Likelihood Framework for Offline Inverse Reinforcement Learning
Siliang Zeng*, Chenliang Li*, Alfredo Garcia, Mingyi Hong
Thirty-seventh Conference on Neural Information Processing Systems (NeurIPS 2023). (Oral: 0.5% [67/12343] )

A Bayesian Approach to Robust Inverse Reinforcement Learning
Ran Wei, Siliang Zeng, Chenliang Li, Alfredo Garcia, Anthony D McDonald, Mingyi Hong
The Conference on Robot Learning (CoRL) 2023.
(A previous version accepted by First Workshop on Theory of Mind in Communicating Agents at ICML 2023)

Maximum-Likelihood Inverse Reinforcement Learning with Finite-Time Guarantees
Siliang Zeng, Chenliang Li, Alfredo Garcia, Mingyi Hong
Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS 2022).
(A previous version accepted by Decision Awareness in Reinforcement Learning Workshop at ICML 2022)

A Stochastic Linearized Augmented Lagrangian Method for Decentralized Bilevel Optimization
Songtao Lu, Siliang Zeng, Xiaodong Cui, Mark S. Squillante, Lior Horesh, Brian Kingsbury, Jia Liu, Mingyi Hong
Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS 2022).
Selected as an Honorable Mention in the IBM Pat Goldberg Memorial competition for best papers.

Learning to Coordinate in Multi-Agent Systems: A Coordinated Actor-Critic Algorithm and Finite-Time Guarantees
Siliang Zeng, Tianyi Chen, Alfredo Garcia, Mingyi Hong
4th Annual Learning for Dynamics & Control Conference (L4DC 2022).

A Near-Optimal Algorithm for Stochastic Bilevel Optimization via Double-Momentum
Prashant Khanduri, Siliang Zeng, Mingyi Hong, Hoi-To Wai, Zhaoran Wang, Zhuoran Yang.
Thirty-fifth Conference on Neural Information Processing Systems (NeurIPS 2021).

Multi-Agent Reinforcement Learning for Adaptive Routing: A Hybrid Method using Eligibility Traces
Siliang Zeng, Xingfei Xu, Yi Chen
2020 IEEE 16th International Conference on Control & Automation (ICCA)

Network-Level System Performance Prediction Using Deep Neural Networks with Cross-Layer Information
Qi Cao, Siliang Zeng, Man-On Pun, Yi Chen
IEEE International Conference on Communications (ICC) 2020