← All masterclasses

Cognitive AI & Spiking Neural Networks

Next-gen bio-inspired architectures — neuromorphic computing for your team.

2–3 days Research engineers, ML engineers, innovation leads

Overview

Transformer-based LLMs dominate today, but the next wave of AI draws from neuroscience. This masterclass introduces your team to bio-inspired computing: spiking neural networks (SNNs), neuromorphic hardware, and cognitive architectures that offer orders-of-magnitude improvements in energy efficiency and real-time processing.

What you’ll learn

  • How biological neurons differ from artificial neurons — and why it matters
  • Practical SNN implementation and training
  • When neuromorphic approaches beat traditional deep learning
  • How to evaluate whether cognitive AI fits your roadmap

Curriculum

Day 1 — From Deep Learning to Cognitive AI

  • Limitations of current architectures: energy, latency, continual learning
  • Biological neural computation: spikes, timing, plasticity
  • Spiking Neural Networks: leaky integrate-and-fire, Izhikevich models
  • Encoding schemes: rate coding, temporal coding, population coding
  • Hands-on: building your first SNN with snnTorch / Norse

Day 2 — Training & Applications

  • Surrogate gradient methods for training SNNs
  • Converting trained ANNs to SNNs
  • Neuromorphic hardware landscape: Intel Loihi, BrainChip Akida, SpiNNaker
  • Applications: edge inference, event-driven vision, robotics, anomaly detection
  • Energy benchmarks: SNN vs transformer for comparable tasks
  • Hands-on: event-driven vision pipeline with DVS camera data

Day 3 — Cognitive Architectures & Your Roadmap

  • Cognitive architectures: ACT-R, SOAR, and modern hybrid approaches
  • Memory systems: episodic, semantic, procedural — beyond key-value stores
  • Combining LLMs with neuromorphic components
  • Evaluating neuromorphic fit for your domain and constraints
  • Building an internal research roadmap
  • Workshop: architecture design session for your use case

Prerequisites

  • Strong Python skills
  • Familiarity with PyTorch or TensorFlow
  • Basic understanding of neural network fundamentals

Outcomes

Your team leaves with working SNN prototypes, a clear understanding of when neuromorphic approaches add value, and a research roadmap tailored to your organisation.

Interested in this masterclass?

Tell me about your team and I'll tailor the programme to your needs.

Book this masterclass