Denis studied Physics at Heidelberg University, where he worked on Neuromorphic Hardware for his bachelor’s thesis. He then continued to study Computational Neuroscience at the Bernstein Center in Berlin. Denis joined the lab for his master’s thesis in 2017 and is currently a PhD student. He is trying to understand how biologically motivated constraints to network architectures can lead to episodic to semantic memory transformations during systems memory consolidation. To this end, he uses methods from machine learning, computational models from neuroscience and mathematical analyses. He is further interested in the olfactory memory system of the Drosophila brain as a model system to investigate mechanisms of memory transfer. Denis is also interested in algorithms for simulating spiking neural networks on graphics processing units and is developing the open source software Brian2CUDA.
Memory transformations in a mechanistic model of systems memory consolidation
In this project we try to understand how episodic to semantic memory transformations can arise in a previously proposed circuit mechanism for systems memory consolidation. We use computational neuroscience models combined with methods from machine learning and mathematical analyses to investigate how architectural and representational constraints can give rise to the transformation of memories. We try to find out how these transformations might on the one hand explain semantisation of memory content and on the other hand control factors such as hippocampus dependency of memories and consolidation speed. To test our proposed circuit mechanism, we investigate memory consolidation of olfactory memories in the Drosophila brain. For this project we collaborate closely with the Owald lab.