About me

Hi! I'm a third-year Ph.D. student at the Electrical and Computer Engineering Department of Carnegie Mellon University. I am fortunate to be advised by Prof. Yuejie Chi. Previously I spent one year at Peking University for a data science master’s program. Before that, I received my B.S. in Mathematics (Honors Program) from Xi'an Jiaotong University. Here is my CV.

Research Interests

I'm interested in machine learning, deep learning, reinforcement learning, optimization and game theory. Specifically, I am interested in developing sample and computationally efficient algorithms for some fundamental machine learning problems.

Selected Publications

• Exploration from a Primal-Dual Lens: Value-Incentivized Actor-Critic Methods for Sample-Efficient Online RL
  NeurIPS 2025
  Tong Yang, Bo Dai, Lin Xiao, Yuejie Chi
  [PDF] [BibTex] @article{yang2025exploration, title={Exploration from a Primal-Dual Lens: Value-Incentivized Actor-Critic Methods for Sample-Efficient Online RL}, author={Yang, Tong and Dai, Bo and Xiao, Lin and Chi, Yuejie}, journal={arXiv preprint arXiv:2506.22401}, year={2025} }

• Multi-head Transformers Provably Learn Symbolic Multi-step Reasoning via Gradient Descent
  NeurIPS 2025
  Tong Yang, Yu Huang, Yingbin Liang, Yuejie Chi
  [PDF] [BibTex] @misc{yang2025multiheadtransformersprovablylearn, title={Multi-head Transformers Provably Learn Symbolic Multi-step Reasoning via Gradient Descent}, author={Tong Yang and Yu Huang and Yingbin Liang and Yuejie Chi}, year={2025}, eprint={2508.08222}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2508.08222}, }

• Incentivize without Bonus: Provably Efficient Model-based Online Multi-agent RL for Markov Games
  ICML 2025
  Tong Yang, Bo Dai, Lin Xiao, Yuejie Chi
  [PDF] [BibTex] @article{yang2025incentivize, title={Incentivize without bonus: Provably efficient model-based online multi-agent RL for Markov games}, author={Yang, Tong and Dai, Bo and Xiao, Lin and Chi, Yuejie}, journal={arXiv preprint arXiv:2502.09780}, year={2025} }

• Faster WIND: Accelerating Iterative Best-of-N Distillation for LLM Alignment
  AISTATS 2025
  Tong Yang, Jincheng Mei, Hanjun Dai, Zixin Wen, Shicong Cen, Dale Schuurmans, Yuejie Chi, Bo Dai
  [PDF] [BibTex] @article{yang2024faster, title={Faster WIND: Accelerating Iterative Best-of-$ N $ Distillation for LLM Alignment}, author={Yang, Tong and Mei, Jincheng and Dai, Hanjun and Wen, Zixin and Cen, Shicong and Schuurmans, Dale and Chi, Yuejie and Dai, Bo}, journal={arXiv preprint arXiv:2410.20727}, year={2024} }

• In-Context Learning with Representations: Contextual Generalization of Trained Transformers
  NeurIPS 2024
  Tong Yang, Yu Huang, Yingbin Liang, Yuejie Chi
  [PDF] [BibTex] @article{yang2024context, title={In-context learning with representations: Contextual generalization of trained transformers}, author={Yang, Tong and Huang, Yu and Liang, Yingbin and Chi, Yuejie}, journal={arXiv preprint arXiv:2408.10147}, year={2024} }

• Federated Natural Policy Gradient and Actor Critic Methods for Multi-task Reinforcement Learning
  NeurIPS 2024
  Tong Yang, Shicong Cen, Yuting Wei, Yuxin Chen, Yuejie Chi
  [PDF] [BibTex] @article{yang2023federated, title={Federated natural policy gradient methods for multi-task reinforcement learning}, author={Yang, Tong and Cen, Shicong and Wei, Yuting and Chen, Yuxin and Chi, Yuejie}, journal={arXiv preprint arXiv:2311.00201}, year={2023} }

• A Primal-Dual Approach to Solving Variational Inequalities with General Constraints
  ICLR 2024
  Tatjana Chavdarova*, Tong Yang*, Matteo Pagliardini, Michael I. Jordan (*equal contribution, order is alphabetical.)
  [PDF] [Poster (OPT@NeurIPS '22)] [BibTex] @article{chavdarova2022acvi, title = {Revisiting the ACVI Method for Constrained Variational Inequalities}, author = {Chavdarova, Tatjana and Pagliardini, Matteo and Yang, Tong and Jordan, Michael I.}, journal= {ArXiv:2210.15659}, year = {2022}, }

• Solving Constrained Variational Inequalities via an Interior Point Method
  ICLR 2023 Spotlight!
  Tong Yang*, Michael I. Jrodan*, Tatjana Chavdarova* (*equal contribution)
  [PDF] [Poster (WiML@ICML '22)] [Code] [BibTex] @article{yang2022acvi, title = {Solving Constrained Variational Inequalities via an Interior Point Method}, author = {Tong Yang and Michael I. Jordan and Tatjana Chavdarova}, journal = {ArXiv:2206.10575}, year = {2022}, }

• Optimization for Amortized Inverse Problems
  ICML 2023
  Tianci Liu*, Tong Yang*, Quan Zhang, Qi Lei (*equal contribution)
  [PDF] [BibTex] @inproceedings{liu2023optimization, title={Optimization for amortized inverse problems}, author={Liu, Tianci and Yang, Tong and Zhang, Quan and Lei, Qi}, booktitle={International Conference on Machine Learning}, pages={22289--22319}, year={2023}, organization={PMLR} }

• Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHF
  ICLR 2025
  Shicong Cen, Jincheng Mei, Katayoon Goshvadi, Hanjun Dai, Tong Yang, Sherry Yang, Dale Schuurmans, Yuejie Chi, Bo Dai
  [PDF] [BibTex] @article{cen2024value, title={Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHF}, author={Cen, Shicong and Mei, Jincheng and Goshvadi, Katayoon and Dai, Hanjun and Yang, Tong and Yang, Sherry and Schuurmans, Dale and Chi, Yuejie and Dai, Bo}, journal={arXiv preprint arXiv:2405.19320}, year={2024} }