About me#

I am a PhD candidate in Informatics at the University of Amsterdam working with Max Welling. My research is on structure-aware machine learning for scientific systems.

I am interested in models that make useful structure explicit: symmetries and graphical structure, with reflections in tensor-network factorizations and Markov kernels.

I was happy to collaborate with and learn from Roberto Bondesan and Kirill Neklyudov.

Current status I expect to complete my PhD in November 2026. In 2025 I worked with Qualcomm AI Research in Amsterdam on efficient structured computation for attention-like layers.

Research direction The short version of my research program is: structure-aware machine learning for quantum systems, connecting probabilistic inference, efficient architectures, and quantum inspired algorithms.

Hot off The Press Myosotis SPIGM workshop poster

An interlude about structure computation, SSM, and attention: Myosotis. I hope to tell more in this line of work later; see the SPIGM workshop poster.

Selected Papers#

Quantum Systems and Structured ML#

  • Myosotis: Structured Computation for Attention-Like Layers (NeurIPS 2025 SPIGM Workshop)
    Evgenii Egorov, Hanno Ackermann, Markus Nagel, Han Cai
    arXiv

  • An Equivariant Machine Learning Decoder for 3D Toric Codes (QCNC 2025)
    Oliver Weissl, Evgenii Egorov
    arXiv

  • The END: An Equivariant Neural Decoder for Quantum Error Correction (ICLR 2023 Workshop on Physics for ML)
    Evgenii Egorov, Roberto Bondesan, Max Welling
    Code arXiv

Probabilistic machine learning, MCMC, and generative models#

  • Ai-sampler: Adversarial Learning of Markov Kernels with Involutive Maps (ICML 2024)
    Evgenii Egorov, Riccardo Valperga, Efstratios Gavves
    Code ICML PMLR arXiv

  • Involutive MCMC: a Unifying Framework (ICML 2020)
    Kirill Neklyudov, Max Welling, Evgenii Egorov, Dmitry Vetrov
    ICML PMLR arXiv

  • MaxEntropy Pursuit Variational Inference (ISNN 2019)
    Evgenii Egorov, Kirill Neklydov, Ruslan Kostoev, Evgeny Burnaev
    Springer arXiv

  • The Implicit Metropolis-Hastings Algorithm (NeurIPS 2019)
    Kirill Neklyudov, Evgenii Egorov, Dmitry Vetrov
    NeurIPS arXiv

  • BooVAE: Boosting Approach for Continual Learning of VAE (NeurIPS 2021)
    Evgenii Egorov, Anna Kuzina, Evgeny Burnaev
    Code NeurIPS arXiv

  • Bayesian Generative Models for Knowledge Transfer in MRI Semantic Segmentation Problems (Frontiers in Neuroscience 2019)
    Anna Kuzina, Evgenii Egorov, Evgeny Burnaev
    DOI

Ongoing Work#

I am thinking about tensor-network contraction, classical shadows, and neural quantum states. Let’s see…

Teaching and Mentoring#

  • Main supervisor, MSc thesis: Structured Neural Shadow Quantum State Tomography, Francien Barkhof, 2025-2026. Grade: 9/10.
  • Main supervisor, MSc thesis: Equivariant Machine Learning Decoder for 3D Toric Codes, Oliver Weissl, 2024-2025. Grade: 8.5/10.
  • Developed materials for the graduate course Deep Bayesian Machine Learning, Skoltech CDISE, autumns 2018-2020.
  • Delivered practical seminars for Bayesian Machine Learning, HSE, HSE Computer Science, 2020, complementing lectures by Dmitry Vetrov.

Experience#

  • PhD Candidate, University of Amsterdam, 2021-present.
  • Research Associate / Research Intern, Qualcomm AI Research, Amsterdam Office, Feb-Aug 2025.
  • Junior Researcher, Skoltech, 2017-2020.

Email#

You can reach me at email dot evgenii dot egorov at gmail dot com. The word email is part of the address.

Personal Forward