Vivek Bharadwaj

I am an EECS graduate student at UC Berkeley, advised by the fantastic pair Jim Demmel and Aydın Buluç. I'm associated with the BeBOP and PASSION groups, and I'm supported by a Department of Energy Computational Science Graduate Fellowship from the Krell Institute.

Interests

Geometric Deep Learning
Sparse Computations
Randomized Sketches
Graph Algorithms

Background

PhD in Computer Science.
UC Berkeley, 2020-Present
BS in CS + Mathematics.
Caltech, 2016-2020

Email: firstname_lastname [at] berkeley.edu

Interests

Geometric Deep Learning
Sparse Computations
Randomized Sketches
Graph Algorithms

Background

PhD in Computer Science.
UC Berkeley, 2020-Present
BS in CS + Mathematics.
Caltech, 2016-2020

Email: firstname_lastname [at] berkeley.edu

news

Jan 26, 2025 01/26/25	Austin Glover and I are thrilled to announce OpenEquivariance, a fast kernel generator for rotation equivariant deep neural networks. These machine learning models rotate their predictions in a compatible way when the input (a point cloud) rotates. Read it here.
Sep 25, 2024 09/25/24	Beheshteh Rakhshan and I, along with Osman Asif Malik and Guillaume Rabusseau, worked on this paper on sketching tensor train core chains. It was just accepted as a poster to NeurIPS 2024.
May 14, 2024 05/14/24	I am thrilled to receive Berkeley's Teaching Effectiveness Award (2023-2024). The prize comes with a $500 check, and the list of winners is online. Read my essay on my blog or on the GSI center website!

recent & upcoming talks

Mar 4, 2025

SIAM CSE 2025 Matrix Computations Minisymposium, Fort Worth, TX

Engineering Fast Kernels for Rotation-Equivariant Deep Neural Networks (Planned)

Nov 1, 2024

Redwood Center for Theoretical Neuroscience Group Meeting, Berkeley, CA

Subspace Embeddings for Khatri-Rao Products (Redux)

Slides Video

May 13, 2024

SIAM Conference on Applied Linear Algebra, Paris, France

Sampling-Based Sketches for Tensor Train Core Chains

Slides

selected publications

* denotes equal contribution. See more papers on the dedicated publications page.

Preprint

An Efficient Sparse Kernel Generator for O(3)-Equivariant Deep Networks

Vivek Bharadwaj*, Austin Glover*, Aydın Buluç, James Demmel

Preprint, 2025

Bib Pdf Code

@misc{openequivariance,
    author = "Bharadwaj, Vivek and Glover, Austin and Buluç, Aydın and Demmel, James",
    title = "An Efficient Sparse Kernel Generator for O(3)-Equivariant Deep Networks",
    year = "2025",
    eprint = "2501.13986",
    archivePrefix = "arXiv",
    primaryClass = "cs.LG",
    url = "https://arxiv.org/abs/2501.13986"
}

NeurIPS

Efficient Leverage Score Sampling for Tensor Train Decomposition

Vivek Bharadwaj*, Beheshteh Rakhshan*, Osman Malik, Guillaume Rabusseau

In Advances in Neural Information Processing Systems, Dec 2024

Bib Pdf Code

@inproceedings{bharadwaj2024efficientleveragescoresampling,
    author = "Bharadwaj, Vivek and Rakhshan, Beheshteh T. and Malik, Osman Asif and Rabusseau, Guillaume",
    title = "Efficient Leverage Score Sampling for Tensor Train Decomposition",
    booktitle = "Advances in Neural Information Processing Systems",
    month = "Dec",
    year = "2024",
    eprint = "2406.02749",
    archivePrefix = "arXiv",
    primaryClass = "cs.DS"
}

NeurIPS

Fast Exact Leverage Score Sampling from Khatri-Rao Products with Applications to Tensor Decomposition

Vivek Bharadwaj, Osman Malik, Riley Murray, Laura Grigori, Aydın Buluç, James Demmel

In Advances in Neural Information Processing Systems, Dec 2023

Bib Pdf Code Poster Slides Web

@inproceedings{bharadwaj23_fast_krp_leverage,
    author = "Bharadwaj, Vivek and Malik, Osman Asif and Murray, Riley and Grigori, Laura and Buluç, Aydın and Demmel, James",
    editor = "Oh, A. and Naumann, T. and Globerson, A. and Saenko, K. and Hardt, M. and Levine, S.",
    booktitle = "Advances in Neural Information Processing Systems",
    pages = "47874--47901",
    publisher = "Curran Associates, Inc.",
    title = "Fast Exact Leverage Score Sampling from Khatri-Rao Products with Applications to Tensor Decomposition",
    volume = "36",
    month = "Dec",
    year = "2023"
}

IPDPS

Distributed-Memory Sparse Kernels for Machine Learning

Vivek Bharadwaj, Aydın Buluç, James Demmel

In 2022 IEEE International Parallel and Distributed Processing Symposium, Jun 2022

Bib Pdf Video Code Poster Slides

@inproceedings{bharadwaj22_distspml,
    author = "Bharadwaj, Vivek and Buluç, Aydın and Demmel, James",
    booktitle = "2022 IEEE International Parallel and Distributed Processing Symposium",
    title = "Distributed-Memory Sparse Kernels for Machine Learning",
    year = "2022",
    volume = "",
    issn = "",
    pages = "47-58",
    keywords = "machine learning algorithms;costs;runtime;collaborative filtering;layout;machine learning;inference algorithms",
    doi = "10.1109/IPDPS53621.2022.00014",
    url = "https://doi.ieeecomputersociety.org/10.1109/IPDPS53621.2022.00014",
    publisher = "IEEE Computer Society",
    address = "Los Alamitos, CA, USA",
    month = "jun"
}