FAQ

Answers to frequently asked questions.

What is NeuronBench?

A no-code way to distribute interactive, biologically realistic neural networks to anyone. A low-code system for building neural networks that guide your students or illustrate your own novel models of neural computation.

Think of NeuronBench whenever you are discussing neural networks and you wish you (and whoever you are talking to) had a real example of the thing you were talking about.

With existing modeling frameworks, you have to distribute installation instructions. With NeuronBench, you send a single URL, and they get a realtime simulation.

Does NeuronBench compete with NEURON, Brian, Arbor, Genesis, etc?

NeuronBench emphasizes interactivity while traditional modeling tools focus on scalability, speed, scriptability and customizability. NeuronBench runs in the browser, while most simulators run on a server or in a cluster. If you are building a model for your research paper and your focus is on the final rendered output, then traditional simulators like NEURON are the more appropriate tool.

How many neurons can I simulate on NeuronBench?

Several very complicated neurons (500 segments), or about 1000 very simple neurons. NeuronBench simulations run directly in the web browser and performance is dependent on how powerful the user's computer is. Simulations can run on the iPad, but not on mobile phones.

Is NeuronBench open source?

Yes! The simulator and the configuration language are developed in the open. The code hosting our website is private for security reasons.

Where can I go for help?

Join the NeuronBench Discord Server to chat with the NeuronBench team and other community members.