Shenghao Yang

Shenghao Yang

Postdoctoral researcher

University of California, Berkeley

International Computer Science Institute


I am a postdoctoral researcher with a joint appointment in the Department of Statistics at UC Berkeley and the International Computer Science Institute, where I am working with Michael Mahoney. My research is at the intersection of machine learning, randomized numerical linear algebra, optimization, and graph algorithms. Previously, I got my PhD in Computer Science at the University of Waterloo.

News

Papers

  • Positional Attention: Expressivity and Learnability of Algorithmic Computation
    Artur Back de Luca*, George Giapitzakis*, Shenghao Yang*, Petar Veličković, Kimon Fountoulakis
    International Conference on Machine Learning (ICML) 2025

    arXiv

  • Using Pre-trained LLMs for Multivariate Time Series Forecasting
    Malcolm L. Wolff, Shenghao Yang, Kari Torkkola, Michael W. Mahoney

    arXiv

  • Polynomial Width is Sufficient for Set Representation with High-dimensional Features
    Peihao Wang, Shenghao Yang, Shu Li, Zhangyang Wang, Pan Li
    International Conference on Learning Representations (ICLR) 2024

    arXiv PDF

  • Local Graph Clustering with Noisy Labels
    Artur Back de Luca, Kimon Fountoulakis, Shenghao Yang
    International Conference on Learning Representations (ICLR) 2024

    arXiv PDF

  • Weighted Flow Diffusion for Local Graph Clustering with Node Attributes: an Algorithm and Statistical Guarantees
    Shenghao Yang, Kimon Fountoulakis
    International Conference on Machine Learning (ICML) 2023

    arXiv PDF Slides Poster Code

  • Graph Attention Retrospective
    Kimon Fountoulakis*, Amit Levi*, Shenghao Yang*, Aseem Baranwal, Aukosh Jagannath
    Journal of Machine Learning Research (JMLR) 2023
    ICLR 2022 Workshop on Anchoring Machine Learning in Classical Algorithmic Theory. Best Paper Award.

    arXiv PDF Code

  • Equivariant Hypergraph Diffusion Neural Operators
    Peihao Wang, Shenghao Yang, Yunyu Liu, Zhangyang Wang, Pan Li
    International Conference on Learning Representations (ICLR) 2023

    arXiv Code

  • On Classification Thresholds for Graph Attention with Edge Features
    Kimon Fountoulakis, Dake He, Silvio Lattanzi, Bryan Perozzi, Anton Tsitsulin, Shenghao Yang

    arXiv Code

  • Local Hyper-Flow Diffusion
    Kimon Fountoulakis, Pan Li, Shenghao Yang
    Neural Information Processing Systems (NeurIPS) 2021

    arXiv PDF Slides (long) Slides (short) Poster Code

  • Targeted Pandemic Containment Through Identifying Local Contact Network Bottlenecks
    Shenghao Yang, Priyabrata Senapati, Di Wang, Chris T. Bauch, Kimon Fountoulakis
    PLOS Computational Biology. 2021

    arXiv DOI Slides Code

  • Parallel and Communication Avoiding Least Angle Regression
    Swapnil Das, James Demmel, Kimon Fountoulakis, Laura Grigori, Michael W. Mahoney, Shenghao Yang
    SIAM Journal on Scientific Computing. 2021

    arXiv DOI

  • p-Norm Flow Diffusion for Local Graph Clustering
    Kimon Fountoulakis, Di Wang, Shenghao Yang
    International Conference on Machine Learning (ICML) 2020

    arXiv PDF Slides Code

  • Split Cuts from Sparse Disjunctions
    Ricardo Fukasawa, Laurent Poirrier, Shenghao Yang
    Mathematical Programming Computation. 2020

    DOI