ICML 2025: Proposed Electrostatic Field Matching (EFM), a physics-inspired method for generative modeling and distribution transfer
ICML 2025: Introduced Inverse Bridge Matching Distillation, accelerating diffusion bridge model inference by 4x–100x with improved generation quality
NeurIPS 2024: Proposed Adversarial Schrödinger Bridge Matching, an efficient iterative procedure for learning Schrödinger Bridges in discrete time
ICLR 2024: Introduced Light Schrödinger Bridge, a novel lightweight solver for the Schrödinger Bridge problem
NeurIPS 2023 (Oral, Top 3%): Proposed Entropic Neural Optimal Transport via Diffusion Processes for computing entropic optimal transport plans from samples
Education
PhD in Math & Physics, 2023 (advisor: Prof. E. Burnaev)
PhD studies in Computer Science, 2018–2022, Skolkovo Institute of Science and Technology (Skoltech)
MSc in Computer Science, 2016–2018, Higher School of Economics (HSE)
Yandex School of Data Analysis, 2013–2016
BSc in Mathematics, 2012–2016, Higher School of Economics (HSE)