Research

2026

1. arXiv

   

   Singular vectors of sparse random rectangular matrices: Transition between delocalized and semilocalized phases

   Ioana Dumitriu and Hai-Xiao Wang \((\alpha\mkern-1mu{-}\mkern-1mu\beta)\)

   arXiv preprint arXiv:2610.22658, 2026

   arXiv Bib

   @article{Dumitriu2026SingularVectors,
     title = {Singular vectors of sparse random rectangular matrices: Transition between delocalized and semilocalized phases},
     author = {Dumitriu, Ioana and Wang, Hai-Xiao},
     journal = {arXiv preprint arXiv:2610.22658},
     year = {2026},
   }

2. arXiv

   

   Concentration of Regularized Sparse Random Matrices: Spectral Edge Bounds via Nonbacktracking Operators

   Hai-Xiao Wang

   arXiv preprint arXiv:2609.26615, 2026

   arXiv Bib

   @article{Wang2026RegularizedSparse,
     title = {Concentration of Regularized Sparse Random Matrices: Spectral Edge Bounds via Nonbacktracking Operators},
     author = {Wang, Hai-Xiao},
     journal = {arXiv preprint arXiv:2609.26615},
     year = {2026},
   }

3. arXiv

   Majority Dynamics on Resampled Sparse Erdős–Rényi Graphs: Gaussian Winner Selection and Pace to Unanimity

   Ioana Dumitriu, Muchen Ju, and Hai-Xiao Wang \((\alpha\mkern-1mu{-}\mkern-1mu\beta)\)

   arXiv preprint arXiv:2608.06159, 2026

   arXiv Bib

   @article{Dumitriu2026ResampledMajority,
     title = {Majority Dynamics on Resampled Sparse Erdős–Rényi Graphs: Gaussian Winner Selection and Pace to Unanimity},
     author = {Dumitriu, Ioana and Ju, Muchen and Wang, Hai-Xiao},
     journal = {arXiv preprint arXiv:2608.06159},
     year = {2026}
   }

4. arXiv

   Majority Dynamics on Assortative Sparse Stochastic Block Models

   Ioana Dumitriu, Muchen Ju, and Hai-Xiao Wang \((\alpha\mkern-1mu{-}\mkern-1mu\beta)\)

   arXiv preprint arXiv:2607.24652, 2026

   arXiv Bib

   @article{Dumitriu2026AssortativeMajority,
     title = {Majority Dynamics on Assortative Sparse Stochastic Block Models},
     author = {Dumitriu, Ioana and Ju, Muchen and Wang, Hai-Xiao},
     journal = {arXiv preprint arXiv:2607.24652},
     year = {2026}
   }

5. arXiv

   Attention Mechanisms Through the Lens of Numerical Methods: Approximation Methods and Alternative Formulations

   Michel Fabrice Serret, Alice Cortinovis, Yijun Dong, Diana Halikias, Anna Ma, Fabio Matti, Deanna Needell, Katherine J. Pearce, Elizaveta Rebrova, Disha Shur, Rudi Smith, Hai-Xiao Wang, and Laura Grigori

   arXiv preprint arXiv:2604.01757, 2026

   arXiv Bib

   @article{Serret2026Attention,
     title = {Attention Mechanisms Through the Lens of Numerical Methods: Approximation Methods and Alternative Formulations},
     author = {Serret, Michel Fabrice and Cortinovis, Alice and Dong, Yijun and Halikias, Diana and Ma, Anna and Matti, Fabio and Needell, Deanna and Pearce, Katherine J. and Rebrova, Elizaveta and Shur, Disha and Smith, Rudi and Wang, Hai-Xiao and Grigori, Laura},
     journal = {arXiv preprint arXiv:2604.01757},
     year = {2026}
   }

2025

1. arXiv

   

   Singular values of sparse random rectangular matrices: Emergence of outliers at criticality

   Ioana Dumitriu, Hai-Xiao Wang, Zhichao Wang, and Yizhe Zhu \((\alpha\mkern-1mu{-}\mkern-1mu\beta)\)

   arXiv preprint arXiv:2508.01456, 2025

   arXiv Bib

   @article{dumitriu2025singular,
     title = {Singular values of sparse random rectangular matrices: Emergence of outliers at criticality},
     author = {Dumitriu, Ioana and Wang, Hai-Xiao and Wang, Zhichao and Zhu, Yizhe},
     journal = {arXiv preprint arXiv:2508.01456},
     year = {2025},
   }

2024

1. ICML 2024

   Optimal Exact Recovery in Semi-Supervised Learning: A Study of Spectral Methods and Graph Convolutional Networks

   Hai-Xiao Wang and Zhichao Wang

   In Proceedings of the 41st International Conference on Machine Learning, 235, pp. 51614–51649, 2024

   arXiv Bib HTML

   @inproceedings{Wang2024OptimalER,
     title = {Optimal Exact Recovery in Semi-Supervised Learning: A Study of Spectral Methods and Graph Convolutional Networks},
     author = {Wang, Hai-Xiao and Wang, Zhichao},
     booktitle = {Proceedings of the 41st International Conference on Machine Learning},
     volume = {235},
     series = {Proceedings of Machine Learning Research},
     pages = {51614--51649},
     publisher = {PMLR},
     url = {https://proceedings.mlr.press/v235/wang24bt.html},
     year = {2024}
   }

2023

1. AoS

   

   Optimal and exact recovery on the general nonuniform Hypergraph Stochastic Block Model

   Ioana Dumitriu and Hai-Xiao Wang \((\alpha\mkern-1mu{-}\mkern-1mu\beta)\)

   The Annals of Statistics, 54(1), pp. 48–73, 2026

   DOI arXiv Bib

   @article{Dumitriu2023OptimalAE,
     title = {Optimal and exact recovery on the general nonuniform Hypergraph Stochastic Block Model},
     author = {Dumitriu, Ioana and Wang, Hai-Xiao},
     journal = {The Annals of Statistics},
     volume = {54},
     number = {1},
     pages = {48--73},
     year = {2026},
     doi = {10.1214/25-AOS2524},
     publisher = {Institute of Mathematical Statistics},
   }

2. arXiv

   Information-Theoretic Limits and Strong Consistency on Binary Non-uniform Hypergraph Stochastic Block Models

   Hai-Xiao Wang

   arXiv preprint arXiv:2306.06845, 2023

   arXiv Bib

   @article{Wang2023ITlimits,
     title = {Information-Theoretic Limits and Strong Consistency on Binary Non-uniform Hypergraph Stochastic Block Models},
     author = {Wang, Hai-Xiao},
     journal = {arXiv preprint arXiv:2306.06845},
     year = {2023}
   }

2021

1. CPC

   Partial recovery and weak consistency in the non-uniform hypergraph stochastic block model

   Ioana Dumitriu, Hai-Xiao Wang, and Yizhe Zhu \((\alpha\mkern-1mu{-}\mkern-1mu\beta)\)

   Combinatorics, Probability and Computing, 34(1), pp. 1–51, 2025

   DOI arXiv Bib

   @article{Dumitriu2025PartialRA,
     title = {Partial recovery and weak consistency in the non-uniform hypergraph stochastic block model},
     author = {Dumitriu, Ioana and Wang, Hai-Xiao and Zhu, Yizhe},
     journal = {Combinatorics, Probability and Computing},
     volume = {34},
     number = {1},
     pages = {1--51},
     year = {2025},
     doi = {10.1017/S0963548324000166},
     publisher = {Cambridge University Press}
   }