1 Foundations of SNNs

Have you wondered how the brain works as a computer? And what it “practically” means... that the brain consists of neurons? What those biological neurons are, and what they do? How can we simulate biological neurons and build neural networks — called Spiking Neural Networks (SNNs) out of them? How learning happens in the brain, and how can we implement it? This topic covers the Computational Neuroscience basics - relevant to the SNNs, and answers all these questions in the following chapters:

What You'll Learn?¶

Here is the list of chapters in this release:

1. What is a Spiking Neuron?: Once you have learned about the biological neurons, the next step is to learn how to mathematically / programmatically simulate them and build Spiking Neuron models. This chapter covers point neuron models.

2. What is Encoding & Decoding?: Real-world data is almost always continuous-valued, and to work with SNNs, we need to represent / convert them to discrete integers. This chapter introduces the concept of Encoding, i.e., learn how to encode continuous values to spikes; and the concept of Decoding, i.e., learn how to decode meaningful information back from discrete spikes.

3. What is plasticity?: Our brain is plastic, but what are the underlying neuroscience principles that facilitate learning in brain? This chapter covers the foundational concepts of neuroplasticity, e.g., Long-Term Potentiation (LTP) and Long-Term Depression (LTD) based on precise spike timings.

Planned for later releases¶

Further chapters are in preparation and will appear in later releases:

• What is a Neuron?: What are biological neurons and how do they function? What role do the glial cells play? What biological characteristics of these neurons and cells should we simulate to build an SNN? This chapter follows through all these fundamental questions.

• Spiking vs Artificial Neurons: how does biological neurons relate to the neurons we are familiar with from ANNs

• Representing data as events: how event-based data is described and manipulated. What information do spikes encode?

• What is an SNN?: This chapter defines a network of spiking neurons and explains how to build Fully Connected SNN and Convolutional SNN architectures from scratch. This chapter demonstrates only the forward (inference) pass. The backward pass / training is covered in the next topic “Training SNNs”.

@incollection{snnbook2026-foundations,
  author    = {Gaurav, Ramashish and Pedersen, Jens Egholm and Bogdan, Petrut},
  title     = {{Foundations of SNNs}},
  booktitle = {{Practical Spiking Neural Networks}},
  publisher = {Open Neuromorphic},
  year      = {2026},
  edition   = {Version 0.8},
  url       = {https://snnbook.net/topics/1_foundations},
}