MesoNet: sigmoid response versus spike threshold

sigmoid response is tempting, but where exactly does the discrete spike happen. threshold adaptation might be the other route—check timing + cost

variants explored: a smooth sigmoid-response neuron and an exponentially adapting threshold. The former smooths the response but changes the relationship between output activity and the discrete events assumed by standard STDP.

ANN and SNN neuron schematic

Neuron response-shape experiment

recorded behavior: sigmoid evaluation was slow; threshold adaptation offered another way to control firing frequency but required tuning. The entry does not contain a quantified speed/accuracy tradeoff.

brian2 pointer: @network_operation for state changes outside the default equation solver; custom event handling for spike-triggered updates.

review questions: which quantity remains discrete in the sigmoid variant, and what triggers the plasticity rule? Compare a hybrid LIF/adaptive-threshold version with the smooth-response version using the same input. Record callback scheduling and timestep so a change in update order is not mistaken for a model effect.

References saved in this journal section

• https://brian2.readthedocs.io/en/latest/resources/tutorials/1-intro-to-brian-neurons.html
• https://brian2.readthedocs.io/en/2.6.0/_modules/brian2/equations/equations.html
• https://brian2.readthedocs.io/en/2.7.1/resources/tutorials/2-intro-to-brian-synapses.html
• https://brian.discourse.group/t/run-error-with-modeling-izhikevich-neuron/579
• https://brian2.readthedocs.io/en/stable/examples/advanced.custom_events.html

Alternative neuron response

Discrete response fluctuations