Exploring Flow-Lenia Universes with a Curiosity-driven AI Scientist: Discovering Diverse Ecosystem Dynamics

arXiv 2025 60.4 method

TLDR

A curiosity-driven AI scientist uses diversity search to discover ecosystem dynamics in Flow-Lenia, a simulated cellular automaton.

Reasoning

The paper presents a novel application of IMGEP for automated discovery in a simulated environment, with strengths in illuminating metric space and scaling analysis. Weaknesses include lack of real-world validation and potential overclaim of generality beyond Flow-Lenia.

Read-first score

Read-first score 60.4, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 74.

Recency 8%

86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=experiment,metric,result

Topical relevance 42%

61.7

Uses existing LLM keyword relevance scores normalized to 0-100. AI scientist,automated scientific discovery,autonomous research agent,automated research,literature review agent,survey generation,automated experimentation,experiment design agent,AI for scientific research,paper writing agent,research automation,scientific discovery agent

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=none

Field roles

FrontierMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 48.

Keyword Scores

AI scientist

automated scientific discovery

AI for scientific research

scientific discovery agent

automated research

automated experimentation

research automation

autonomous research agent

experiment design agent

literature review agent

survey generation

paper writing agent

Deep Analysis

Innovations

Curiosity-driven AI scientist method for discovering system-level dynamics in Flow-Lenia
Adaptation of Intrinsically Motivated Goal Exploration Processes (IMGEPs) to large environments of interacting patterns using simulation-wide metrics (evolutionary activity, compression ratio, multi-scale matter distribution)
Use of diversity search archive to scaffold subsequent scaling experiments, enabling an iterative experiment design loop with an interactive exploration tool
Discovery of macro-scale organization with no analogue at the base scale through scaling study

Methodology

The authors adapt IMGEPs to Flow-Lenia, a continuous cellular automaton with mass conservation, using simulation-wide metrics to explore large environments of interacting patterns. Two exploration experiments target ecosystem-level dynamics and matter movement through obstacles, comparing IMGEP against random search. A scaling study across six spatial scales and seven time horizons is then conducted using the resulting archive.

Key Results

IMGEP illuminates significantly more of the metric space than random search, reveals self-organized behaviors qualitatively resembling biological phenomena, and uncovers macro-scale organization absent at the base scale.

Limitations

Demonstrated only on Flow-Lenia; generalizability to other parameterizable complex systems is not yet shown.

Resources

Paper