Neuronal algorithms


Research center

45 rue d’Ulm
75230 Paris
Marc Mézard


Ecole Normale Supérieure


Institut de Biologie de l'ENS IBENS
U1024 UMR 8197
Labex : "mémolife" (IBENS, CdF, ESPCI), Idex : PSL*




Bouvier, G., Clopath, C., Bimbard, C., Nadal, J.-P., Brunel, N., Hakim, V. and Barbour, B. Cerebellar learning using perturbations. bioRxiv 053785; doi:


Blot, A. and Barbour, B. Ultra-rapid axon-axon ephaptic inhibition of cerebellar Purkinje cells by the pinceau. Nature neuroscience, 2014, Vol. 17, pp. 289-295

Ostojic, S., Szapiro, G., Schwartz, E., Barbour, B., Brunel, N. and Hakim, V. Neuronal morphology generates high-frequency firing resonance. The Journal of neuroscience, 2015, Vol. 35, pp. 7056-7068

Ly, R., Bouvier, G., Szapiro, G., Prosser, H. M., Randall, A. D., Kano, M., Sakimura, K., Isope, P., Barbour, B. and Feltz, A.  Contribution of postsynaptic T-type calcium channels to parallel fibre-Purkinje cell synaptic responses. The Journal of physiology, 2016, Vol. 594, pp. 915-936

Bouvier, G., Higgins, D., Spolidoro, M., Carrel, D., Mathieu, B., Léna, C., Dieudonné, S., Barbour, B., Brunel, N. and Casado, M. Burst-Dependent Bidirectional Plasticity in the Cerebellum Is Driven by Presynaptic NMDA Receptors. Cell reports, 2016, Vol. 15, pp. 104-116

Fields of research

Neurophysiology / systems neuroscience

Research Theme

Our group aims to understand the operation and function of one particular brain region, the cerebellum, basing our research upon a multidisciplinary characterization of the neural networks it contains. _* The cerebellum is involved in the learning and execution of coordinated movements. As a brain structure in which to study the representation and transformation of information, it offers several significant advantages : it has a simple and well-described cellular architecture, we know something about both its (sensory) inputs and (motor) outputs, and several well-understood model behaviors are very strongly linked to the cerebellum.

We study cerebellar function using three principal approaches :

  • The in vitro characterization of the network, neurons and synapses of the cerebellum, using patch-clamp recording and imaging in slices.
  • In vivo recordings of cerebellar activity during behavior, using tetrodes to monitor the behavior of multiple neurons simultaneously.
  • Theoretical analysis and numerical modeling (often in collaboration with theoretical physicists).

Team members

Clément Lena
Mathieu Tihy
Annick Ayon
Anne Feltz
Daniela Popa
German Szapiro

Lab rotation

A new algorithm for cerebellar learning

Team leader: 



April 2, 2018 - June 29, 2018

Application deadline: 

June 29, 2018


~ April-June 2018


We have recently proposed a stochastic gradient descent algorithm for cerebellar learning and identified a possible cellular implementation in the olivo-cerebellar circuitry ( We are now testing our hypothesis by combining in vivo electrophysiological recordings during behviour, theoretical analysis and slice electrophysiology. Rotation students would be welcome to join the project at any of these study levels.


Ecole Normale Supérieure - IBENS - 46 rue d'Ulm 75005 Paris - +33 1 44 32 37 36 -