Gennadii Filatov

Second-year M.Sc. student in Big Data and Machine Learning at ITMO University.

IMG_5309.PNG

I am a second-year M.Sc. student in Big Data and Machine Learning at ITMO University, supervised by Irina Deeva in the AIST Laboratory. Irina has since moved to HSE University, where I continue this research under her supervision.

My research focuses on synthetic tabular data: how to calibrate generative models so that synthetic data reliably preserves model rankings, enabling trustworthy evaluation of machine learning systems. I am also interested in generative modeling more broadly, including diffusion models and flow matching for structured and sequential data. More broadly, I am interested in generative models and probabilistic inference.

In summer 2026, I was selected for the SMILES International Summer School on Machine Learning, where I contributed to a project on few-step text generation using categorical flow maps.

Previously, I completed a B.Sc. in Physics (Nuclear and Particle Physics) at Peter the Great St. Petersburg Polytechnic University, where I worked on experimental data analysis at the PRES experiment at SPbPU.

Honors & Awards

  • Erasmus+ Mobility Grant (2026-2027) — competitive EU-funded academic exchange at Alexandru Ioan Cuza University of Iasi, Romania, awarded through a selective institutional process.
  • Enhanced State Academic Scholarship (2022–2025) — top 5% of SPbPU students
  • Student of the Year (2024) — Peter the Great St. Petersburg Polytechnic University
  • IAPS Worldwide Grant (2024) — for participation in ICPS 2024

Research Interests

  • Synthetic tabular data: generation, calibration, and evaluation
  • Generative models (diffusion models, flow matching)
  • Probabilistic inference and uncertainty quantification

selected publications

  1. AISTATS WS
    When Synthetic Data Is Enough: Calibration for Tabular Model Ranking
    Gennadii Filatov and Irina Deeva
    In Towards Trustworthy Predictions: Theory and Applications of Calibration for Modern AI, 2026