Hieu N. Nguyen

I'm often amazed by the vastness of the ocean.

👋 Hi! I’m Hieu. I am a first-year Ph.D. student at Penn State NLP, where I’m advised by Prof. Rui Zhang. I study the strengths and limitations of learning algorithms through carefully designed and controlled experimentation, with an emphasis on failure modes and robustness. I believe this understanding can help us:

Develop more efficient learning algorithms (WWW ‘25, ICLR ‘26) and more robust systems (ICLR ‘25, NeurIPS ‘25, ACL ‘23),
Build systems capable of reasoning and producing creative solutions (arXiv ‘26), and
Apply them to accelerate scientific discovery (ICML ‘25) and understand our world.

Outside of research and work, I enjoy learning violin 🎻 as a hobby in my free time. I also like playing sports 🏸🏀🥏, so feel free to say hi or invite me if you see me around. These activities help me generate new ideas 💡 and stay creative in my research.

Also See my thoughts on research and my generated tokens (blog posts). You can find my academic résumé here (last updated: April 2026).

Last updated: June 1, 2026

selected publications

For an up-to-date list of my research papers, please see my Google Scholar profile. * denotes equal contribution.

Why Do Reasoning Models Lose Coverage? The Role of Data and Forks in the Road Preprint

Ngoc-Hieu Nguyen^*, Parshin Shojaee^*, Phuc Minh Nguyen, and 4 more authors

2026

Bib Blog Code

@article{hieu2026reasoningforks,
  title = {Why Do Reasoning Models Lose Coverage? The Role of Data and Forks in the Road},
  author = {Nguyen, Ngoc-Hieu and Shojaee, Parshin and Nguyen, Phuc Minh and Zhang, Nan and Reddy, Chandan K and Doan, Khoa D and Zhang, Rui},
  year = {2026},
  eprint = {2605.17026},
  archiveprefix = {arXiv},
  primaryclass = {cs.LG},
  url = {https://arxiv.org/abs/2605.17026},
}

Mitigating Reward Over-optimization in Direct Alignment Algorithms with Importance Sampling NeurIPS

Nguyen Minh Phuc, Ngoc-Hieu Nguyen, Duy Minh Ho Nguyen, and 5 more authors

In The Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025

Bib

@inproceedings{phuc2025mitigating,
  title = {Mitigating Reward Over-optimization in Direct Alignment Algorithms with Importance Sampling},
  author = {Phuc, Nguyen Minh and Nguyen, Ngoc-Hieu and Nguyen, Duy Minh Ho and Liu, Anji and Mai, An and Nguyen, Binh T. and Sonntag, Daniel and Doan, Khoa D},
  booktitle = {The Thirty-ninth Annual Conference on Neural Information Processing Systems},
  year = {2025},
  url = {https://openreview.net/forum?id=ltPRj2nthL},
}

LLM-SRBench: A New Benchmark for Scientific Equation Discovery with Large Language Models ICML Oral

Parshin Shojaee^*, Ngoc-Hieu Nguyen^*, Kazem Meidani, and 3 more authors

In Forty-second International Conference on Machine Learning, 2025

Awarded Bib Code Slides

Oral presentation: 1% acceptance

@inproceedings{shojaee2025llmsrbench,
  title = {{LLM}-{SRB}ench: A New Benchmark for Scientific Equation Discovery with Large Language Models},
  author = {Shojaee, Parshin and Nguyen, Ngoc-Hieu and Meidani, Kazem and Farimani, Amir Barati and Doan, Khoa D and Reddy, Chandan K.},
  booktitle = {Forty-second International Conference on Machine Learning},
  year = {2025},
  url = {https://openreview.net/forum?id=SyQPiZJVWY},
}

Wicked Oddities: Selectively Poisoning for Effective Clean-Label Backdoor Attacks ICLR

Nguyen Hung-Quang, Ngoc-Hieu Nguyen, The-Anh Ta, and 4 more authors

In The Thirteenth International Conference on Learning Representations, 2025

Bib

@inproceedings{hung-quang2025wicked,
  title = {Wicked Oddities: Selectively Poisoning for Effective Clean-Label Backdoor Attacks},
  author = {Hung-Quang, Nguyen and Nguyen, Ngoc-Hieu and Ta, The-Anh and Nguyen-Tang, Thanh and Wong, Kok-Seng and Thanh-Tung, Hoang and Doan, Khoa D},
  booktitle = {The Thirteenth International Conference on Learning Representations},
  year = {2025},
  url = {https://openreview.net/forum?id=1Z3C49JQVf},
}

Towards Efficient Communication and Secure Federated Recommendation System via Low-rank Training WWW Oral

Ngoc-Hieu Nguyen, Tuan-Anh Nguyen, Tuan Nguyen, and 3 more authors

In Proceedings of the ACM Web Conference 2024, Singapore, Singapore, 2024

Awarded DOI Bib Code Slides

Oral presentation

@inproceedings{hieu2024colr,
  author = {Nguyen, Ngoc-Hieu and Nguyen, Tuan-Anh and Nguyen, Tuan and Hoang, Vu Tien and Le, Dung D. and Wong, Kok-Seng},
  title = {Towards Efficient Communication and Secure Federated Recommendation System via Low-rank Training},
  year = {2024},
  isbn = {9798400701719},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  url = {https://doi.org/10.1145/3589334.3645702},
  doi = {10.1145/3589334.3645702},
  booktitle = {Proceedings of the ACM Web Conference 2024},
  pages = {3940–3951},
  numpages = {12},
  keywords = {communication efficiency, federated learning, recommendation system},
  location = {Singapore, Singapore},
  series = {WWW '24},
}

The images are generated from the paper’s title and abstract using Nano Banana. Can you infer the abstract from the visuals?