The following tutorials provide a step-by-step introduction to some off the shelf models. To be able to follow this tutorial we will assume you have already compiled Auryn and you know how to run your own Auryn simulations (see quick start and compile and run).
- Tutorial 1: Simulate a single AdEx neuron and record spikes and membrane potentials.
- Tutorial 2: Build the Vogels & Abbott balanced network model.
- Tutorial 3: Create and simulate a network model with inhibitory plasticity.
- Tutorial 4: SuperSpike: Supervised learning in spiking networks (coming soon).
Advanced techniques: Extending the Auryn model corpus
As you have seen Auryn already comes with a variety of neuronal and synaptic plasticity models as well as devices to interact with and to record from your network simulations (see manual). However, in many cases you will want to create your own models. Auryn is written in a modular way which greatly simplifies the process. In this section you will learn step-by-step how.
- Writing your own plasticity model. This is a simple walk-through for the logic behind plastic updates and what methods are called where and when. It sketches in simple terms what needs to be done to implement a new custom synapse model in Auryn.
- Multiple synaptic state variables. This example aims at creating a plastic connection object in which the the actual weight change is the low-pass filtered output of meta-variable which is influenced by STDP.
tutorials/start.txt · Last modified: 2018/05/30 09:21 by zenke