February 23, 2017, 2:00pm Title: "How is space-time represented in the brain:
The physics of fuzzy memory"
Abstract:
Neural Networks perform parallel computations in a fuzzy
(meaning imprecise) fashion to represent space, time and memory,
which in turn can be flexibly used for decision-making with regard to
potential future events.
I will give an overview of how cognitive-neuroscience researchers have attempted to theorize this.
Further, I will present a neural-net model where the connection weights are
derived as an inverse-Laplace transform operation.
I will try to convince you that the computational simplicity of this
neural-network architecture provides all the flexibility needed to store and
extract space, time and memory representations.
February 24, 2017, 4:00pm Title: "Can Torsion induced cosmology avert
gravitational singularities at the classical level?"
Abstract:
I will present a modification to General Relativity (GR) induced by
torsion in an extra-dimension (can be interpreted as Kaluza-Klein
theory with torsion). The only physical constraint imposed on this
theory is that any motion through the extra-dimension should not affect
the observable 4D motion so as to keep the extra-dimension invisible.
Interestingly, this theory exactly mimics GR at the kinematic level;
hence it would reproduce the basic results like light-bending and
perihelion precession of mercury. However, at the dynamical level, the
field equations get modified from GR and reveals singularity free
solutions at the cosmological scales. In particular, the model admits
solutions like extremely compact nonsingular objects (which for all
observational purposes might appear as black holes), and oscillatory
universe that can bounce back from an arbitrarily small size without
hitting a singularity. This reveals a possibility that we could
potentially eradicate gravitational singularities within classical
theories of gravity by exploiting torsion.