Conclusion to Other Complex Neuron Models and Features

• Really complex models, channels and compartments, can really help understand particular neurons.
• If you can't build a model of something, you don't understand it.
• The models are, necessarily, incomplete.
• I want to mention two papers: "Cell assembly formation and structure in a piriform cortex model" Traub, Tu and Whittington (2022), and "Reconstruction and simulation of neocortical microcircuitry" (2015) by Markram and a lot of other folks.
• The Traub paper uses a few thousand compartmental neurons to simulate a model of cell assemblies. The Markram paper describes a really large model with dozens of types of neurons that is meant to accurately represent the behaviour of a small area of cortex.
• These are both great but take a lot of time to simulate.
• Moreover, the Traub paper acknowledges that it's not using enough neurons.
• How well can simluated neurons model cognition?
• Reading this week is to look over Trappenberg 2.4 again.
• Reading for next week is Trappenberg Chapter 4, but particularly Section 4.2.