|
How can we learn to recognize stimuli ? And how can we still remember them decades later?| old_uid | 9300 |
|---|
| title | How can we learn to recognize stimuli ? And how can we still remember them decades later? |
|---|
| start_date | 2010/11/23 |
|---|
| schedule | 13h |
|---|
| online | no |
|---|
| location_info | bâtiment BSHM, etage 1, salle D32 |
|---|
| summary | The brain is capable of storing memories for stimuli that have only been seen relatively briefly. Here, I will argue that Spike-Time Dependent Plasticity (STDP) is a key mechanism underlying this ability, especially when combined with the idea that information in sensory pathways is encoded in the temporal patterning of firing in neurons. Simulation studies have demonstrated that when a pattern of afferent activity is repeatedly presented, neurons will become selective to that pattern, and progressively learn to respond to the earliest part of the pattern. I will discuss a recent study using auditory noise patterns that provide experimental support for this proposal (Agus et al, 2010, Neuron, 66, 610). I will also show how this sort of STDP based selectivity mechanism can potentially explain how we are able to remember stimuli that we have not experienced for decades. But I will note that the proposal depends on two controversial ideas : (1) the suggestion that neurons can become extremely selective (effectively becoming "grandmother cells"), and (2) the possibility that substantial numbers of cortical neurons might have effectively zero spontaneous activity - a sort of cortical "dark matter". |
|---|
| responsibles | Pascalis, Kandel |
|---|
| |
|