Nai-Chieh Huang

About me

I am a first-year Ph.D. student in the Machine Learning Department at Carnegie Mellon University, where I am fortunate to be advised by Prof. Max Simchowitz. My research interests lie broadly in reinforcement learning, optimization, generative modeling, and control. Specifically, I study RL dynamics, understanding the fundamental principles of RL that govern modern AI systems, to enable stable, predictive, and continually improving training pipelines.

Previously, I received my B.S. in Computer Science from National Yang Ming Chiao Tung University in Taiwan. As an undergraduate, I conducted research under the supervision of Prof. Ping-Chun Hsieh, focusing on convergence guarantees in reinforcement learning.

News

Dec 06, 2025	Our new paper Much Ado About Noising is out! We present a large-scale empirical study of whether generative control policies genuinely improve behavior cloning, or if it’s all much ado about noising.
Jul 11, 2025	Excited to join CMU MLD as a PhD student! Can’t wait to visit Pittsburgh and meet everyone soon!

Selected Publications

Much Ado About Noising: Dispelling the Myths of Generative Robotic Control

Chaoyi Pan, Giri Anantharaman, Nai-Chieh Huang, Claire Jin, Daniel Pfrommer, and 6 more authors

arXiv preprint arXiv:2512.01809, 2025

arXiv Bib PDF Website

@article{pan2025much,
  title = {Much Ado About Noising: Dispelling the Myths of Generative Robotic Control},
  author = {Pan, Chaoyi and Anantharaman, Giri and Huang, Nai-Chieh and Jin, Claire and Pfrommer, Daniel and Yuan, Chenyang and Permenter, Frank and Qu, Guannan and Boffi, Nicholas and Shi, Guanya and Simchowitz, Max},
  journal = {arXiv preprint arXiv:2512.01809},
  year = {2025},
}

Accelerated Policy Gradient: On the Convergence Rates of the Nesterov Momentum for Reinforcement Learning

Yen-Ju Chen^*, Nai-Chieh Huang^*, Ching-pei Lee, and Ping-Chun Hsieh

In International Conference on Machine Learning (ICML), 2024

arXiv Bib PDF

@inproceedings{chen2024accelerated,
  title = {Accelerated Policy Gradient: On the Convergence Rates of the Nesterov Momentum for Reinforcement Learning},
  author = {Chen, Yen-Ju and Huang, Nai-Chieh and Lee, Ching-pei and Hsieh, Ping-Chun},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2024},
}

PPO-Clip Attains Global Optimality: Towards Deeper Understandings of Clipping

Nai-Chieh Huang, Ping-Chun Hsieh, Kuo-Hao Ho, and I-Chen Wu

In Proceedings of the AAAI Conference on Artificial Intelligence, 2024

arXiv Bib PDF

@inproceedings{huang2024ppo,
  title = {PPO-Clip Attains Global Optimality: Towards Deeper Understandings of Clipping},
  author = {Huang, Nai-Chieh and Hsieh, Ping-Chun and Ho, Kuo-Hao and Wu, I-Chen},
  booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2024},
}