A world of tiny creatures, driven by simple rules, yet creating complex societies.
Emergant is an ant colony simulator built with PyGame, where reinforcement learning and neuro-evolution drives the emergence of lifelike ant behaviors. Watch as virtual ants forage for food, defend their nests, and adapt their strategies over time—all without explicit scripting of their behaviors!
- 🌱 Reinforcement Learning – Ants from the same colony are all collected by one hivemind, and learn to bring food back to their colony using a system of rewards and penalties.
- 🧠 Neuroevolutionary Algorithm - The neural network structure of each hivemind evolves from generation to generation through natural selection.
- 🎮 PyGame Visualization – The simulation runs in real-time with an interactive graphical interface. There is also a neural network visualization utility to get deeper insights into the weights of the networks and to increase explainability.
(Insert GIFs or screenshots of the simulation in action!)
- Clone the repository:
git clone https://github.com/matanitah/emergant.git cd emergant - Install dependencies:
pip install -r requirements.txt
- Start the simulation:
python src/main.py
Reach out to me on Linkedin at https://www.linkedin.com/in/matan-itah/ if you are interested in becoming a collaborator.
MIT License – Free to use and modify!