Neuroscience and Artificial Intelligence

Welcome

PhD - Postdoctoral Researcher
Institute of Neural Information Processing, Center for Molecular Neurobiology Hamburg (ZMNH), University Medical Center Hamburg-Eppendorf (UKE), Hamburg

I study how information and memory is learned, maintained, consolidated and processed in biological and artificial networks. More specifically, my research aims to model these neuronal mechanisms at different scales from cellular (e.g. spiking neural networks) to population levels (e.g. phase oscillators, mean-field models).

In particular, I investigate the role of inhibitory inter-neurons in these processes. From another perspective, I also study how slow oscillations affect functionnal connectivity.

My overall goal is both to gain a better understanding of the biological mechanisms that govern the brain and to incorporate certain neural principles into artificial intelligence models.

Alumni of:

Neurocybernetics team, ETIS lab, CY Cergy Paris University, Cergy.
Computational Neuroscience Group, Center for Brain and Cognition, Pompeu Fabra University, Barcelona.

Research areas

Artificial neural networks
Spiking neurons
Coupled oscillators
Mean-field models
Computational neuroscience
Functionnal and structural connectivity
Learning and plasticity
Memory
Oscillations
Multimodal learning and processing
Inhibition
Synchronization

Last news

October 9, 2026 - Attendance to Network Neuroscience Symposium, Hamburg, Germany

September 8, 2026 - On-site review of the proposed Collaborative Research Center (CRC) 1794, Hamburg, Germany

Title: Multi-scale Neural Dynamics in Healthy and Disordered Cognition

August 31, 2026 - Attendance to an invited talk of Chris Summerfield, Hamburg, Germany

March 25, 2026 - CNRS qualifications, France

CNRS qualifications for the position of Maître de conférences (MCF) / Associate Professor in Sections 27 (Computer Science) and 69 (Neuroscience).

December 2, 2025 - Attendance to the "Hamburg neuro talks", Hamburg, Germany

November 5-6, 2025 - SNUFA 2025, Online.

Contributed talk entitled "Emergence and maintenance of modularity in neural networks with Hebbian and anti-Hebbian inhibitory STDP".

September 11, 2025 - Attendance to the "Hamburg neuro talks", Hamburg, Germany

September 4, 2025 - Attendance to the ZMNH retreat, Hamburg, Germany

June 23-27, 2025 - XLV Dynamics Days Europe, Thessaloniki, Greece

Talk for the minisymposium: Neural Dynamics Across Spatiotemporal Scales: Models, Learning Processes, Computational Tools, and Clinical Applications, entitled "Coherent and rich neuronal dynamics supported by different inhibition mechanisms".

April 22, 2025 - Publication of a journal paper in PLoS Computational Biology

R. Bergoin, A. Torcini, G. Deco, M. Quoy, G. Zamora-López. Emergence and maintenance of modularity in neural networks with Hebbian and anti-Hebbian inhibitory STDP. PLoS Comput Biol 21(4): e1012973, 2025.

November 5-16, 2024 - Attendance to the workshop SNUFA 2024, Online.

October 15-17, 2024 - Attendance to the workshop "Introduction to Recurrent Neural Networks and their Applications" from the Center for Biomedical AI at the University Medical Center Hamburg-Eppendorf (UKE).

September 7, 2024 - Publication of a conference paper in From Animals to Animats: 17th International Conference on the Simulation of Adaptive Behavior (SAB 2024)

L. L’Haridon, R. Bergoin, B.S. Bal, M. Abdelwahed and L. Cañamero. The Emergence of a Complex Representation of Touch Through Interaction with a Robot. In: O. Brock, J. Krichmar (eds) From Animals to Animats 17. SAB 2024. Lecture Notes in Computer Science(), vol 14993. Springer, Cham.

Consolidation and maintenance of modularity

Sustainable memory learning: Spontaneous recall of the learned items in the network dynamics at rest promotes the consolidation of associated structural clusters, and thus the long-term maintenance of memories.

Memory storage: The number of inhibitory neurons (i.e. pairs of Hebbian and anti-Hebbian inhbitory neurons) present in the network is correlated with the number of structural clusters that can be formed and stabilized over time, and therefore with the memory storage capacity of the network.

Integration of bimodal information

This spiking neural network is capable of recognizing and generating previously learned audio-visual digits in a multimodal scenario.

In particular, sub-classes of inhibitory neuron have distinct roles in this processing.

Astrocytes calcium enhances slow oscillations

During external stimulation, mean-field astrocyte populations respond to precise range of slow oscillations, affecting neuronal populations activity through calcium modulation.

During spontaneous activity, this astrocyte calcium modulation also enhances slow oscillations across bain regions.