MesoNet: simulator selection and learning resources

Brian2 once again. need the equations + custom events exposed, the convenient abstractions are kind of in the way here

options considered: Brian2, PyNN, BindsNET and snnTorch. The selection criterion in the notebook is access to neuron equations, custom events and synaptic updates—not only ready-made classification examples.

decision at this stage: continue with Brian2 because higher-level abstractions in the alternatives could obstruct the intended dynamics experiments. PyNN remains a possible later route for cross-simulator work. Compatibility/interchange claims in these old notes should be verified against the actual tool versions before porting.

study reference: the EPFL neuronal-dynamics lectures cover neuron-model choices and their assumptions. Use the recorded simulator comparison as a project-specific decision, not a claim that the alternatives cannot support custom research.

review question: what is the smallest example demonstrating each needed capability—custom reset, separate event pathways, plastic synapses, and recording internal state?

References saved in this journal section

• https://open-neuromorphic.org/neuromorphic-computing/software/snn-frameworks/)
• https://bindsnet-docs.readthedocs.io/)
• https://snntorch.readthedocs.io/)
• https://briansimulator.org/)
• https://lcnwww.epfl.ch/gerstner/NeuronalDynamics-MOOCall.html

Brian2 model and console