Neuroevolution—the optimization of neural networks using evolutionary computation—has been an active and steadily growing research area since the 1990s. Traditional machine learning methods require explicit targets or gradients, but neuroevolution excels where these are unavailable—such as in reinforcement learning, robotic control, game-playing, and sequential decision-making tasks in complex environments.
More recently, neuroevolution has expanded beyond evolving behavior policies to:
- automatically designing deep neural architectures,
- modeling the evolution of biological intelligence,
- and optimizing neural networks for hardware-aware deployment.
This community site extends the ideas introduced in the book
Neuroevolution: Harnessing Creativity in AI Agent Design
and serves as a collaborative hub to help researchers, educators, and practitioners share resources—from labs and libraries to benchmarks, tutorials, and papers.
Pull requests are welcome!
- Provide an up-to-date overview of the neuroevolution community
- Highlight research groups, libraries, benchmarks, applications, and papers
- Help newcomers find starting points
- Support instructors, students, and practitioners
- Encourage open-source collaboration and reproducible research
Contributions are managed through GitHub pull requests. To contribute:
- Fork this repository.
- Edit the relevant section below to add your entry (see templates).
- Ensure entries are:
- Publicly accessible (no broken links)
- Brief and informative (1–3 sentences)
- Neutral in tone
- Submit a pull request with a short summary of your changes.
- Research Groups
- Software & Libraries
- Benchmarks & Environments
- Tutorials, Courses & Lectures
- Workshops & Special Issues
- Community & Discussion
- UT Austin – Neural Networks Research Group – Focus on evolving deep neural architectures and controllers for reinforcement learning tasks. Website
- IT University of Copenhagen – Creative AI lab – Specializes in neuroevolution and evolving collective systems. Website
- **Group Name** (Institution, Country)
Brief summary of the group's research focus.
Links: [website](URL) · [GitHub](URL) · [Twitter](URL) (optional)
- neat-python (Python, BSD) – A NEAT implementation in pure Python. GitHub
- EvoTorch (Python, Apache-2.0) – A PyTorch-based evolutionary optimization library. Docs · GitHub
- SharpNEAT (C#, MIT) – .NET-based NEAT implementation. GitHub
- TensorNEAT (Python/JAX, GPL) – GPU-accelerated NEAT for JAX. GitHub
- QDax (Python, MIT) – JAX implementation of MAP-Elites and QD algorithms. GitHub
- **Project Name** (Language, License) – One-line description of the tool or framework.
Links: [GitHub](URL) · [Docs](URL) · [Paper](URL) (optional)
- OpenAI Gym / Gymnasium – Standard RL benchmarks used for neuroevolution.
- Google Brax – Differentiable physics simulator for fast NE experiments.
- Evolution Gym – Modular environments for evolving soft robot morphology and control. GitHub
- Neural MMO – Large-scale multi-agent world for evolving behaviors. GitHub
- **Benchmark Name** – One-sentence description of what the benchmark evaluates.
Links: [website](URL) · [GitHub](URL) · [Paper](URL) (optional)
- GECCO 2024 Tutorial: Evolution of Neural Networks – Tutorial by Risto Miikkulainen.
- Neuroevolution: Harnessing Creativity in AI Agent Design (Book) – Website
- CS378 Neuroevolution – Course at UT Austin. Course page
- **Title** (Year, Venue/Institution) – Very brief description of who it's for and what it covers.
Links: [slides](URL) · [video](URL) · [code](URL)
- **Workshop Title** (Conference/Journal, Year–Year) – Brief scope of the workshop.
Links: [website](URL) · [call for papers](URL) · [proceedings](URL)
- SIGEVO Mailing List – Announcements and CFPs. Site
- **Community Name** – Short description of who it's for.
Links: [invite link](URL) · [website](URL)
This page is maintained by the neuroevolution community. Please contribute or suggest improvements via pull request!