Nikolay Malkin

Nikolay Malkin is a researcher at the University of Edinburgh specializing in machine learning, generative modeling, and computational biology. His work focuses on simulation-free methods for stochastic dynamics and optimal transport, particularly in applications like single-cell dynamics and protein structure prediction.

• Research Interests: Flow-based generative models, Schrödinger bridges, diffusion models, and their applications in high-dimensional biological data analysis.
• Key Contributions: Development of [SF]²M for cell dynamics modeling, generalized conditional flow matching (CFM) for improved generative tasks, and subtrajectory balance (SubTB) for GFlowNet stability.
• Collaborations: Active collaboration with Yoshua Bengio, Alexander Tong, and other researchers in advancing flow matching and generative modeling techniques.
• Publishing Trends: Recent papers emphasize protein backbone generation, tree-structured Schrödinger bridges, and velocity-growth dynamics in single-cell populations.