cv

Doctoral research fellow

Simula Research Laboratory

PhD project focusing on the neural mechanisms of navigation and memory. The research involves developing models to replicate hippocampal neural dynamics, particularly through generative predictive coding in neurons such as place and grid cells during navigational tasks.

Head engineer in scientific computing

Norwegian University of Life Sciences (NMBU)

Project position within the Human Brain Project. Primarily work on development of the cutting-edge brain simulation tool NEST (NEural Simulation Tool) in an international team. My focus is on implementing and optimizing methods for brain-scale constructions on peta- and exascale computers.

Research assistant

Center for Computing in Science Education, University of Oslo

Developed course material for an introductory course in Python programming and computational modeling in biosciences.

Teaching assistant

University of Oslo

Teaching assistant in courses on programming and computational modeling at both an introductory and intermediate bachelor level. In addition, I was a teaching assistant on machine learning and data analysis at master's level.

M.Sc. in Computational Science

University of Oslo

• Computational neuroscience
• Computational physics
• Machine learning
• Bayesian data analysis
• High-performance computing

B.Sc. in Physics

University of Oslo

• Classical physics
• Modern physics
• Scientific computing
• Biophysics

A simplified model of NMDA-receptor-mediated dynamics in leaky integrate-and-fire neurons

Journal of Computational Neuroscience

A model for NMDA-receptor-mediated synaptic currents in leaky integrate-and-fire neurons, first proposed by Wang (J Neurosci, 1999), has been widely studied in computational neuroscience. The model features a fast rise in the NMDA conductance upon spikes in a pre-synaptic neuron followed by a slow decay. In a general implementation of this model which allows for arbitrary network connectivity and delay distributions, the summed NMDA current from all neurons in a pre-synaptic population cannot be simulated in aggregated form. Simulating each synapse separately is prohibitively slow for all but small networks, which has largely limited the use of the model to fully connected networks with identical delays, for which an efficient simulation scheme exists. We propose an approximation to the original model that can be efficiently simulated for arbitrary network connectivity and delay distributions. Our results demonstrate that the approximation incurs minimal error and preserves network dynamics. We further use the approximate model to explore binary decision making in sparsely coupled networks.

Metamodelling of a two-population spiking neural network

PLoS Computational Biology

Biophysical simulations of neuronal networks exhibit complex dynamics, and the parameterization of the neurons and synapses in the network shape the model's behavior. These simulations are generally computationally intensive, especially when fitting models to experimental data. This study explores various metamodelling techniques, which are faster data-driven approximations, to model the population spiking activities and local field potential generated by a two-population recurrent network model. By inverting these metamodels, the study reliably identifies parameter combinations that produce specific simulation outputs.