Hoang-Son Nguyen | Oregon State University

About

My current research focuses on AI systems that provably learn a truthful, robust, causal understanding of the real world. These capabilities will enable AI applications in scientific discovery, sequential decision making, counterfactual reasoning, and so on.

I’m working towards my M.S. degree at Oregon State University, with Xiao Fu. Previously, I received my B.Eng. from The Chinese University of Hong Kong in 2024, working with Hoi-To Wai.

Outside of work, my hobbies are cooking, falling in love with Asian cinema, and photographing my life with close friends whenever I can.

Feel free to ping me an email for a coffee chat!

Questions

How can we train an agent to curiously interact with the environment, acquire intervention data, and learn a causal world model?
How do training data influence the internal representations of LLM/VLM?
Connection between graph learning and deep representation learning?
How to recursively compress an agent’s actions towards a hierarchical structure?

News

[Jan. 2026] Our work on robustness of graph learning algorithms is accepted at ICASSP 2026!
[Sep. 2025] Diverse Influence Component Analysis is accepted at NeurIPS 2025!
[Sep. 2025] A preprint on graph learning is available here, showing that graph topology learning from smooth signals is implicitly robust to missing nodes.
[Apr. 2025] Our contribution on graph learning under partial observation was accepted at the Graph Signal Processing Workshop 2025. The poster is here.
[Sep. 2024] I gave a talk on low-pass graph signal processing at Faculty of Data Science and AI, National Economics University in Hanoi, Vietnam; the slides are here.
[Jul. 2024] Our work on low-pass graph signal detection under partial observations was awarded Best Student Paper Award at IEEE SAM 2024!

Publications

NeurIPS

Diverse Influence Component Analysis: A Geometric Approach to Nonlinear Mixture Identifiability

Hoang-Son Nguyen, Xiao Fu

Conference on Neural Information Processing Systems (NeurIPS), 2025.

PDF Code Poster

ICASSP

Learning Graph from Smooth Signals under Partial Observation: A Robustness Analysis

Hoang-Son Nguyen, Hoi-To Wai

IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2026.

PDF Code

SAM

On Detecting Low-pass Graph Signals under Partial Observations

Hoang-Son Nguyen, Hoi-To Wai

IEEE Sensor Array and Multichannel Signal Processing Workshop (SAM), 2024.

PDF Poster Best Student Paper Award

ICASSP

On the Stability of Low Pass Graph Filter with a Large Number of Edge Rewires

Hoang-Son Nguyen, Yiran He, Hoi-To Wai

IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2022.

PDF Poster Slides

Awards

Best Student Paper Award at SAM, 2024.
Charles K. Kao Research Exchange Scholarship, 2023.
CUHK CSE Outstanding Academic Performance, 2021.
CUHK Admission Scholarship, 2019.