Repository for Group Evolving Agents (GEA), a new paradigm for open-ended self-improvements, which treats a group of agents as the fundamental evolutionary unit, enabling explicit experience sharing and reuse within the group throughout evolution.
# API keys, add to ~/.bashrc
export OPENAI_API_KEY='...'
export ANTHROPIC_API_KEY='...'# Verify that Docker is properly configured in your environment.
docker run hello-world
# If a permission error occurs, add the user to the Docker group
sudo usermod -aG docker $USER
newgrp docker# Install dependencies
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
# Optional: for running analysis
sudo apt-get install graphviz graphviz-dev
pip install -r requirements_dev.txt# Clone SWE-bench
cd swe_bench
git clone https://github.com/princeton-nlp/SWE-bench.git
cd SWE-bench
git checkout dc4c087c2b9e4cefebf2e3d201d27e36
pip install -e .
cd ../../
# Prepare Polyglot
# Make sure git is properly configured in your environment with username and email
python -m polyglot.prepare_polyglot_datasetpython GEA_outer.pyanalysis/scripts used for plotting and analysisinitial/SWE-bench logs and performance of the initial agentinitial_polyglot/Polyglot logs and performance of the initial agentswe_bench/code needed for SWE-bench evaluationpolyglot/code needed for Polyglot evaluationprompts/prompts used for foundation modelstools/tools available to the foundation modelscoding_agent.pymain implementation of the initial coding agentGEA_outer.pyentry point for running the GEA algorithm
This codebase is built upon the Darwin Gödel Machine (DGM). We sincerely thank the authors for their inspiring and impactful work.
The evaluation framework implementations are based on the SWE-bench and polyglot-benchmark repositories.
If you find this project useful, please consider citing:
@article{weng2026group,
title={Group-Evolving Agents: Open-Ended Self-Improvement via Experience Sharing},
author={Weng, Zhaotian and Antoniades, Antonis and Nathani, Deepak and Zhang, Zhen and Pu, Xiao and Wang, Xin Eric},
journal={arXiv preprint arXiv:2602.04837},
year={2026}
}