A Framework for Application-Oriented Design of Large-Scale Neural Networks

Tools for simulations of neural networks exist aplenty. They range from simulators for detailed multi-compartment neurons, over packages for precise reconstruction of small biological networks, to simulators for large-scale networks with stochastic connectivity properties. However, no frameworks for constructing large-scale, dedicated networks exist. Based on the design principles used for our previous work, we introduce a C++ framework which is specifically tailored to simplify the construction of large networks with specific cognitive functionalities.