Lecturer: Prof. Haim Sompolinsky
Affiliation: Racah Institute of Physics,
The Hebrew University of Jerusalem
Abstract:
The tight coupling between excitatory
and inhibitory signals has been long
regarded as a key factor in the
dynamics of brain circuits, and
disruption of this interplay has been
implicated in many brain disorders. In
my lecture, I will survey the theoretical
advances in understanding of this
phenomenon, starting from the
development, twenty years ago, of the
Balanced Network Theory. This theory
has shown that simple neuronal
circuits with strong synaptic
interactions dynamically settle into a
state in which the excitatory currents
are roughly balanced by the inhibitory
ones. The emergent chaotic dynamics
provided a natural explanation of the
ubiquitous irregular, asynchronous
firing patterns observed in cortical
circuits. I will discuss our current
understanding of the dynamic and
computational consequences of
balanced networks.
Anatomical studies in recent years
have allowed us to construct realistic
models of local cortical circuits,
comprising several ten thousand
neurons. This has provided for the first
time a unique opportunity for testing
and refining abstract theoretical
principles such as balanced dynamics
in realistic circuits.
Finally, I will describe a recent work
that indicates that excitation-inhibition
balance emerges not only from
network dynamics but also from
synaptic learning in neural networks of
associative memory.