My research asks how learning algorithms develop structure, select solutions, and shape the behavior of trained models. I combine optimization, differential equations, probability, and geometry with algorithm design and experiments.
When do dominant components emerge, and how do representations organize during training? I study low-rank learning, early stopping, and neural collapse.
How do parameterization, initialization, and update rules determine which data-fitting solution training selects? I derive selection criteria and convergence guarantees.
What does learned structure reveal about unfamiliar inputs and generated outputs? My work includes out-of-distribution detection and semantic accessibility in diffusion models.
How to induce regularization in linear models: A guide to reparametrizing gradient flow.
Hung-Hsu Chou, Johannes Maly, Dominik Stöger.
Advances in Computational Mathematics, 52, article 75, 2026.
Edem Boahen, Simone Brugiapaglia, Hung-Hsu Chou, Mark Iwen, Felix Krahmer.
Journal of Complexity, 96:102065, 2026.
Robust implicit regularization via weight normalization.
Hung-Hsu Chou, Holger Rauhut, Rachel Ward.
Information and Inference, 13(3):iaae022, 2024.
Gradient descent for deep matrix factorization: dynamics and implicit bias towards low rank.
Hung-Hsu Chou, Carsten Gieshoff, Johannes Maly, Holger Rauhut.
Applied and Computational Harmonic Analysis, 68:101595, 2024.
More is less: inducing sparsity via overparameterization.
Hung-Hsu Chou, Johannes Maly, Holger Rauhut.
Information and Inference, 12(3):1437–1460, 2023.
Overparameterization and generalization error: Weighted trigonometric interpolation.
Yuege Xie, Hung-Hsu Chou, Holger Rauhut, Rachel Ward.
SIAM Journal on Mathematics of Data Science, 4(2):885–908, 2022.
Accuracy, efficiency and optimization of signal fragmentation.
Russel E. Caflisch, Hung-Hsu Chou, Jonathan W. Siegel.
Multiscale Modeling & Simulation, 18(2):737–757, 2020.
Optimal correlation estimators for quantized signals.
Michael D. Johnson, Hung-Hsu Chou, Carl R. Gwinn.
The Astrophysical Journal, 765:135, 2013.
Conflicting Biases at the Edge of Stability: Norm versus Sharpness Regularization.
Maria Matveev, Vit Fojtik, Hung-Hsu Chou, Gitta Kutyniok, Johannes Maly.
ICML, 2026.
GradPCA: Leveraging NTK Alignment for Reliable Out-of-Distribution Detection.
Mariia Seleznova, Hung-Hsu Chou, Claudio Mayrink Verdun, Gitta Kutyniok.
ICLR, 2026.
Get rid of your constraints and reparametrize: A study in NNLS and implicit bias.
Hung-Hsu Chou, Johannes Maly, Claudio Mayrink Verdun, Bernardo Freitas Paulo da Costa, Heudson Mirandola.
AISTATS, 2025.
Neural (Tangent Kernel) Collapse.
Mariia Seleznova, Dana Weitzner, Raja Giryes, Gitta Kutyniok, Hung-Hsu Chou.
NeurIPS, 2023.
Stability of Low-Rank Implicit Regularization in Perturbed Deep Matrix Factorization.
Jingzhe Wang, Hung-Hsu Chou.
arXiv:2605.28613, 2026.
From Seeds to Semantics: Measuring Semantic Accessibility in Deterministic Diffusion Models.
Kuntian Chen, Wei Wei, Yizhou Zeng, Sophie Langer, Mariia Seleznova, Hung-Hsu Chou.
arXiv:2602.06155, 2026.
Compressive Sensing—Introduction and Relations to Deep Learning.
Hung-Hsu Chou, Johannes Maly, Holger Rauhut.
Lecture notes for the CIME Doctoral School; arXiv:2608.24211, 2026.
Efficient frame projection of amplitude-modulated signals.
Russel E. Caflisch, Hung-Hsu Chou, Jonathan W. Siegel.
UCLA CAM report, 2018.