This repository contains the companion website for the paper "Curiosity-driven exploration of Flow Lenia universes: Bootstrapping open-ended evolution", presented at the ALife Conference 2025.
Link to the webpage: http://developmentalsystems.org/Exploring-Flow-Lenia-Universes/
The project presents a systematic approach for discovering diverse ecosystem-level emergent properties in complex systems using Intrinsically Motivated Goal Exploration Processes (IMGEP). This methodology is extended to explore evolutionary dynamics at the system level, using Flow-Lenia—a continuous cellular automaton with mass conservation and parameter localization—as a test case. The work demonstrates how goal-directed exploration can efficiently navigate vast parameter spaces to identify conditions promoting open-ended evolution.
Key aspects include:
- Modifications to Flow-Lenia to better support multi-species evolution.
- Evaluation of systems using simulation-wide metrics: evolutionary activity, compression-based complexity, and multi-scale entropy.
- An interactive exploration tool to complement automated discovery, creating an effective human-AI collaborative workflow for scientific investigation.
- Extension of IMGEP as a systematic discovery tool for ecosystem-level emergent properties in complex systems, applied to uncover diverse evolutionary dynamics in Flow-Lenia.
- A multi-metric approach to evaluating evolutionary dynamics, combining evolutionary activity, complexity measures, and multi-scale entropy analysis.
- Discovery of diverse evolutionary phenomena suggesting ecological interactions.
An interactive visualization tool is available on the website to explore Flow-Lenia's evolutionary dynamics. Users can examine parameter configurations, emergent behaviors, and evolutionary metrics discovered by IMGEP.