Alexander van Meegen

This paper develops a theory of interpretable compositional computation using low-rank recurrent neural networks, identifying shared latent components and their task-dependent usage.

The paper develops a multi-scale spiking network model of human cortex to explore structure-dynamics relationships using multimodal data and validates it against electrophysiological and fMRI data.

The paper shows neuron densities in mammalian cerebral cortex follow a lognormal distribution across and within cortical areas, suggesting a new organizational principle.

The paper introduces an open-source Python toolbox for analyzing spiking neuronal networks using mean-field methods, enabling estimation of network properties without simulation.