Venue: BBS
Registration: evento.renater.fr
Organized in the frame of Fjola Hyseni’s thesis defense (Decembre 1st – Broca Center)
With (see details below)
15:00 – 15:40 – Albert COMPTE – IDIBAPS, Barcelona
15:40 – 16:20 – Hervé ROUAULT – Université Aix Marseille
16:20 – 17:00 – Arthur LEBLOIS – IMN, Bordeaux
Coffee break
17:20 – 18:00 – Daniela VALLENTIN – MPI, Seewiesen
18:00 – 18:40 – M. Alexander PITTI – Cergy-Paris Université
Albert COMPTE – IDIBAPS
Title: Prefrontal mechanisms involved in learning distractor-resistant working memory in a dual task
Abstract: Working memory (WM) relies on temporarily storing stimulus features in the activity of neuronal populations. To preserve these dynamics from distraction it has been proposed that pre and post-distraction population activity decomposes into orthogonal subspaces. Using calcium imaging data from the prelimbic (PrL) and anterior cingulate (ACC) cortices I will provide evidence that WM orthogonalization emerges as mice learn to perform an olfactory dual task, which combines an outer Delayed Paired-Association task (DPA) with an inner Go-NoGo task. As mice learned the task, the projection of pre-distraction population activity onto the low-dimensional subspace that encodes the distractor odors gradually vanished. A computational model links learning to (1) the orthogonalization of sample and distractor WM subspaces and (2) the orthogonalization of each subspace with irrelevant inputs, which we validated in the experimental data using optogenetic manipulations. In the model, learning drives the network from a double-well attractor toward a more continuous ring attractor regime. We tested signatures for this dynamical evolution in the experimental data by estimating the energy landscape of the dynamics on a one-dimensional ring. In sum, our study defines network dynamics underlying the process of learning to shield WM representations from distracting tasks.
Hervé ROUAULT – Université Aix Marseille
More details soon
Arthur LEBLOIS – IMN, Bordeaux
Title : Dopamine-dependent plasticity and attractor dynamics cooperate for movement adaptation in the basal ganglia-thalamo-cortical loop
Abstract : The basal ganglia (BG) are subcortical nuclei known for their implication in motor control, sensorimotor integration and procedural learning. The BG shape activity patterns in their thalamocortical target structures to optimize actions. In particular, plasticity in BG output implements corrections in behavior to maximize reward and correct errors during learning. The BG are however not necessary for the production of already-learned movement. Based on current anatomical, physiological and behavioral evidence, we have built a model for the generation, learning and adaptation of reaching movement in the BG-thalamo-cortical loop. In this model, we aim at studying the interplay between the attractor dynamics in the motor cortex and the dynamics of the closed-loop BG-thalamo-cortical network during reward-driven learning enabled by DA-dependent long-term plasticity at the cortico-striatal synapses. We propose that the BG-thalamo-cortical network can shape the motor output based on a dual mechanism involving the rich dynamics of this closed-loop network and the classical RL mechanisms relying on dopamine-dependent cortico-striatal synaptic plasticity.
Daniela VALLENTIN – MPI, Seewiesen
Title: Neural mechanisms of vocal interactions in songbirds
Abstract: During conversations we rapidly switch between listening and speaking which often requires withholding or delaying our speech in order to hear others and avoid overlap. The ability of vocal turn-taking is exhibited by non-linguistic species as well, however the neural circuit mechanisms that enable us to regulate the precise timing of our vocalizations during interactions are largely unknown.We address this issue by studying zebra finches that coordinate their calls during vocal interactions By performing intracellular recordings and pharmacological manipulation in the premotor nucleus HVC we found that inhibition regulates the coordination of social contact calls. To further study more complex vocal interactions we also study the singing behavior of nightingales. Male nightingales learn over 100 different songs which are used in order to attract mates or defend territories. We investigated auditory-induced vocal plasticity in interacting nightingales and discovered that nightingales rapidly and accurately imitated the pitch of pitch-controlled whistle playbacks. This finding highlights their capability of directly transforming an auditory input to a matching vocal response.
M. Alexander PITTI – Cergy-Paris Université
Title : Model of structure learning in Broca area