Small Spiking Systems

For fun and learning.

↓ Slide 2 / 6

Simulations

Professional:

Multi-agent systems.

Hobby:

Emergent swarming of boids.
Boltzmann Machines.
An RSNN that trains in the browser.
RSNN with SNNTorch.
SNN with Numpy only.
Pathfinding with the Free Energy Principle.
Ising Model Visualisations.
Classifying spectrum from quantum dots.

↓ Slide 3 / 6

Goal: Oscillator networks with a task.

✅ Small Oscillator XOR network, not self-organising
✅ Larger self-organising oscillating network.
❌ Larger self-organising oscillating network with task.

↓ Slide 4 / 6

Oscillator-XOR Network

↓ Slide 5 / 6

Larger Oscillator-only Network

Firing rate set as target for homeostasis.
Conductance-based synapses.
No plasticity.
80 E / 20 I neurons.

15.1Hz average firing rate of E neurons.
86.3Hz average firing rate of I neurons.
27.0Hz oscillation frequency of E neurons.

Next steps:

Integrate a task.
Robustness checks.

Ping Spike Raster Plot

Spectrogram of E Neurons

↓ Slide 6 / 6

Next steps

Goal: Understand SOTA of this question.

Goal: Larger self-organising PING network with task.

How?

Main network performs a task itself.
Or oscillatory subnetwork generates timing signals for a task network.

If 2., how many signals passed between subnetworks, then can mock the main network for testing and robustness checks?

Other options:

More self organisation, local learning of weights.
Add training, grad-desc, or teacher forcing.

Slide 1 / 6