Synaptic

Hybrid neural simulation and digital logic framework built around:

• A configurable Hodgkin-Huxley neuron model
• Event-driven synaptic network orchestration
• NAND-only digital architecture primitives and composed circuits

Features

• Hodgkin-Huxley neuron dynamics from scratch (hodgkinhuxley.py)
• Synapse and network orchestration for multi-neuron circuits (connection.py)
• NAND-only fabric with derived gates, adder, comparator, and ALU (nand_architectures.py)
• Demo runner with plotting and CLI options (main.py)

Repository Layout

.
├── connection.py
├── hodgkinhuxley.py
├── main.py
├── nand_architectures.py
├── requirements.txt
├── CONTRIBUTING.md
├── LICENSE
└── docs
    ├── API.md
    └── ARCHITECTURE.md

Quick Start

1. Create and activate a virtual environment.
2. Install dependencies:

pip install -r requirements.txt

3. Run all demos:

python main.py

CLI Usage

Run HH-only demo:

python main.py --mode hh

Run NAND-only demos:

python main.py --mode nand

Run without opening plot windows:

python main.py --no-show

Save generated figures:

python main.py --save-prefix output/demo

This writes:

• output/demo_hh.png
• output/demo_nand_scaling.png

Core Concepts

• HH coincidence circuit: Two upstream neurons (A and B) converge onto C. Tunable synaptic weights and pulse timing let you explore coincidence-driven firing behavior.
• NAND universality: Complex digital behavior (adder/comparator/ALU) is composed from NAND primitives only.

Documentation

• API reference: docs/API.md
• Architecture notes: docs/ARCHITECTURE.md

Development

See CONTRIBUTING.md for style, checks, and workflow.

License

MIT License. See LICENSE.