🧠 SynCo: A Modular Spiking Synthetic Cortex

Created by Aliyu Lawan Halliru (2025)
License: MIT

SynCo is a biologically inspired, spiking neural network that mimics real brain dynamics using:

⚡ Spiking neurons (LIF, Adaptive LIF)
🧬 Local synaptic learning (STDP, Hebbian)
🧠 Modular cognitive architecture (Relay, Memory, Comparator, Feedback)
🧪 Reinforcement-ready outputs with multi-task switching
🔁 Lifelong learning via task replay and local plasticity

This model bridges neuroscience and artificial general intelligence, enabling realistic, interpretable, and continual learning from sparse feedback and spiking dynamics.

📦 Files Included

SynCo_Synthetic_Cortex_Demo.ipynb: Notebook with full training demo
final_modular_brain_agent_with_spikes_and_plasticity.py: Full model code
README.md: This file

🧪 Example Output

Step 04 | Task: binary         | Loss: 0.0123 | acc: 1.00
Step 12 | Task: classification | Loss: 1.2391 | acc: 0.88
Step 19 | Task: regression     | Loss: 0.5214

SynCo adapts its weights in real time using only local neuron activity — no backpropagation required.

🧠 Use Cases

Neuroscience-inspired AI modeling
Continual learning agents
Synthetic cortex simulation
Educational use in bio-AI and neural computation

✨ Credits

Created by Aliyu Lawan Halliru, Nigerian independent AI researcher.
Project aims to make synthetic neuroscience accessible to the world.

📜 License

MIT License — free to use and adapt.