top of page


Stability and learning in excitatory synapses by nonlinear inhibitory plasticity

New paper out! In this PLoS CB publication, we reveal that a novel form of inhibitory plasticity, which depends non-linearly on the firing rate of the neuron, is a sufficient mechanism to homeostatically stabilize excitatory rate and plasticity dynamics while still enabling flexible learning upon inhibition.

Our author summary:

"An important task the brain needs to solve is the so-called ‘stability-flexibility problem’. On the one hand, any representation in the brain, for example a long-lasting memory, has to be stable for a long time. On the other hand, new representations need to be flexibly learned at any time. Learning and memory formation are implemented through the plasticity of synaptic connections, which describe how the activity in neurons is translated into changes of synaptic strength between these neurons. We propose a novel form of synaptic plasticity at synapses from inhibitory to excitatory neurons as a mechanism to stabilize learned representations, while a gating signal triggers the learning of new representations. We identify the dominance of inhibition over excitation and a nonlinear dependence of inhibitory plasticity on the postsynaptic firing rate as important aspects of our newly proposed plasticity mechanism. Our computational model allows us to uncover the underlying mechanism behind various experimental findings related to synaptic plasticity and sensory perturbations, and we formulate multiple experimentally-testable predictions."

Miehl C., Gjorgjieva J. (2022). “Stability and learning in excitatory synapses by nonlinear inhibitory plasticity.” PLoS Comput Biol, 18(12):e1010682.

Twitter thread:


Commenting has been turned off.
bottom of page