Grigory Bartosh

Ph.D. Candidate in ML at UvA

I am a last-year Ph.D. student in the Amsterdam Machine Learning Lab (AMLab) at the University of Amsterdam, supervised by Christian A. Naesseth.

Previously, I completed my Master’s in Data Science at the Higher School of Economics, under the supervision of Dmitry Vetrov. I was also a researcher at JetBrains Research and completed research internships at Microsoft Research (Cambridge) and in Tim Salimans’ group at Google DeepMind (Amsterdam).

I work in the field of Deep Learning, with research interests focused on Generative Models, particularly Diffusion Models.

Selected publications

ICLR

Forward-Learned Discrete Diffusion: Learning how to noise to denoise faster

Grigory Bartosh, Teodora Pandeva, Sushrut Karmalkar, and 1 more author

The 14th International Conference on Learning Representations (ICLR), 2026

Bib

@article{bartosh2026forward,
  title = {Forward-Learned Discrete Diffusion: Learning how to noise to denoise faster},
  author = {Bartosh, Grigory and Pandeva, Teodora and Karmalkar, Sushrut and Zazo, Javier},
  journal = {The 14th International Conference on Learning Representations (ICLR)},
  year = {2026},
}

ICML

SDE Matching: Scalable and Simulation-Free Training of Latent Stochastic Differential Equations

Grigory Bartosh, Dmitry Vetrov, and Christian A Naesseth

The 42st International Conference on Machine Learning (ICML), 2025

arXiv Bib

@article{bartosh2025sde,
  title = {SDE Matching: Scalable and Simulation-Free Training of Latent Stochastic Differential Equations},
  author = {Bartosh, Grigory and Vetrov, Dmitry and Naesseth, Christian A},
  journal = {The 42st International Conference on Machine Learning (ICML)},
  year = {2025},
}

NeurIPS

Neural Flow Diffusion Models: Learnable Forward Process for Improved Diffusion Modelling

Grigory Bartosh, Dmitry Vetrov, and Christian A Naesseth

Advances in Neural Information Processing Systems, 2024

arXiv Bib

@article{bartosh2024neural,
  title = {Neural Flow Diffusion Models: Learnable Forward Process for Improved Diffusion Modelling},
  author = {Bartosh, Grigory and Vetrov, Dmitry and Naesseth, Christian A},
  journal = {Advances in Neural Information Processing Systems},
  volume = {37},
  year = {2024},
}

NeurIPS

Variational Flow Matching for Graph Generation

Floor Eijkelboom^*, Grigory Bartosh^*, Christian Andersson Naesseth, and 2 more authors

Advances in Neural Information Processing Systems, 2024

arXiv Bib

@article{eijkelboom2024variational,
  title = {Variational Flow Matching for Graph Generation},
  author = {Eijkelboom, Floor and Bartosh, Grigory and Naesseth, Christian Andersson and Welling, Max and van de Meent, Jan-Willem},
  journal = {Advances in Neural Information Processing Systems},
  volume = {37},
  year = {2024},
}

NeurIPS

Equivariant Neural Diffusion for Molecule Generation

François RJ Cornet, Grigory Bartosh, Mikkel N Schmidt, and 1 more author

Advances in Neural Information Processing Systems, 2024

Bib HTML

@article{cornet2024equivariant,
  title = {Equivariant Neural Diffusion for Molecule Generation},
  author = {Cornet, Fran{\c{c}}ois RJ and Bartosh, Grigory and Schmidt, Mikkel N and Naesseth, Christian A},
  journal = {Advances in Neural Information Processing Systems},
  volume = {37},
  year = {2024},
}

ICML

Neural Diffusion Models

Grigory Bartosh, Dmitry Vetrov, and Christian A Naesseth

The 41st International Conference on Machine Learning (ICML), 2024

arXiv Bib

@article{bartosh2023neural,
  title = {Neural Diffusion Models},
  author = {Bartosh, Grigory and Vetrov, Dmitry and Naesseth, Christian A},
  journal = {The 41st International Conference on Machine Learning (ICML)},
  year = {2024},
}