ML Engineer · Google

Haiyang Huang

Research on efficient
machine learning systems.

Efficient ML systems · Scalable inference · Interpretable learning

Explore my research Google Scholar GitHub LinkedIn

Haiyang Huang — Efficient systems. Interpretable learning.

About me

I’m an ML Engineer at Google. I graduated with a PhD in Computer Science from Duke University, where I was advised by Cynthia Rudin and Benjamin C. Lee.

During my PhD, I studied efficient machine learning systems and scalable inference, including memory use and inference efficiency for mixture-of-experts models. My doctoral research also covered interpretable machine learning and dimensionality reduction for data visualization.

Before Duke, I earned a bachelor’s degree in Mathematics and Computer Science at the University of Michigan, where I worked with Jenna Wiens.

Efficient ML systems

Improving large-scale inference through dynamic gating, expert buffering, and load balancing for mixture-of-experts models.

Explore the work

Interpretable learning

Understanding visual concepts and building models whose decisions we can inspect.

Data visualization

Revealing local and global structure in high-dimensional data.

Selected research

All publications

AISTATS

The Rashomon Effect for Visualizing High-Dimensional Data

Yiyang Sun , Haiyang Huang, Gaurav Rajesh Parikh , and 1 more author

In Proceedings of the 29th International Conference on Artificial Intelligence and Statistics , 2026

arXiv Bib HTML

@inproceedings{sun2026rashomon,
  title = {The Rashomon Effect for Visualizing High-Dimensional Data},
  author = {Sun, Yiyang and Huang, Haiyang and Parikh, Gaurav Rajesh and Rudin, Cynthia},
  booktitle = {Proceedings of the 29th International Conference on Artificial Intelligence and Statistics},
  volume = {300},
  pages = {3160--3168},
  year = {2026},
}

AAAI

Dimension Reduction with Locally Adjusted Graphs

Yingfan Wang , Yiyang Sun , Haiyang Huang, and 1 more author

In Proceedings of the AAAI Conference on Artificial Intelligence , 2025

arXiv Bib HTML

@inproceedings{wang2025localmap,
  title = {Dimension Reduction with Locally Adjusted Graphs},
  author = {Wang, Yingfan and Sun, Yiyang and Huang, Haiyang and Rudin, Cynthia},
  booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence},
  volume = {39},
  number = {20},
  pages = {21357--21365},
  year = {2025},
  doi = {10.1609/aaai.v39i20.35436},
}

NeurIPS

Toward Efficient Inference for Mixture of Experts

Haiyang Huang, Newsha Ardalani , Anna Sun , and 5 more authors

In Advances in Neural Information Processing Systems , 2024

Bib HTML Code

@inproceedings{huangtoward,
  title = {Toward Efficient Inference for Mixture of Experts},
  author = {Huang, Haiyang and Ardalani, Newsha and Sun, Anna and Ke, Liu and Bhosale, Shruti and Lee, Hsien-Hsin S and Wu, Carole-Jean and Lee, Benjamin},
  booktitle = {Advances in Neural Information Processing Systems},
  year = {2024},
}

NeurIPS

Navigating the Effect of Parametrization for Dimensionality Reduction

Haiyang Huang, Yingfan Wang , and Cynthia Rudin

In Advances in Neural Information Processing Systems , 2024

arXiv Bib HTML Code

@inproceedings{huangnavigating,
  title = {Navigating the Effect of Parametrization for Dimensionality Reduction},
  author = {Huang, Haiyang and Wang, Yingfan and Rudin, Cynthia},
  booktitle = {Advances in Neural Information Processing Systems},
  year = {2024},
}

JMLR

Understanding How Dimension Reduction Tools Work: An Empirical Approach to Deciphering t-SNE, UMAP, TriMap, and PaCMAP for Data Visualization

Yingfan Wang , Haiyang Huang, Cynthia Rudin , and 1 more author

Journal of Machine Learning Research, 2021

Bib HTML PDF Code

@article{wang2021understanding,
  title = {Understanding How Dimension Reduction Tools Work: An Empirical Approach to Deciphering t-SNE, UMAP, TriMap, and PaCMAP for Data Visualization},
  author = {Wang, Yingfan and Huang, Haiyang and Rudin, Cynthia and Shaposhnik, Yaron},
  journal = {Journal of Machine Learning Research},
  volume = {22},
  pages = {1--73},
  year = {2021},
}

Updates

Apr 01, 2026	The Rashomon Effect for Visualizing High-Dimensional Data is accepted at AISTATS 2026.
Apr 11, 2025	Our LocalMAP paper is published in the AAAI 2025 proceedings.
Nov 10, 2024	Two of my papers were accepted at NeurIPS 2024: efficient MoE inference and parametric dimensionality reduction.
May 06, 2022	I am thrilled to join Facebook AI Research (FAIR) as a research intern starting at the middle of May!

Let’s connect.

Reach out on LinkedIn for research conversations and collaborations.

Connect on LinkedIn ORCID DBLP