analiza aiqbench
This research adapts Intrinsically Motivated Goal Exploration Processes to Flow-Lenia, demonstrating that curiosity-driven agents can efficiently map complex metric spaces and uncover macro-scale organizational patterns absent at the base level. By using low-cost diversity searches as a scaffold for expensive experiments, the study establishes an iterative loop that keeps human scientists in the loop while scaling across multiple spatial and temporal horizons. The approach offers a principled framework for studying bottom-up collective behavior in other parameterizable complex systems beyond cellular automata.
Exploring Flow-Lenia Universes with a Curiosity-driven AI Scientist: Discovering Diverse Ecosystem Dynamics
arXiv:2505.15998v4 Announce Type: replace Abstract: We present a curiosity-driven AI scientist method for discovering system-level dynamics in Flow-Lenia, a continuous cellular automaton (CA) with mass conservation and parameter localization. Building on prior work that uses diversity search in Lenia to find individual self-organized patterns, we adapt Intrinsically Motivated Goal Exploration Processes (IMGEPs) to large environments of interacting patterns, using simulation-wide metrics such as evolutionary activity, compression ratio, and multi-scale matter distribution. We apply IMGEP in two exploration experiments: one targeting ecosystem-level dynamics, the other matter movement through obstacle-laden environments. In both, IMGEP illuminates significantly more of the metric space than random search and reveals self-organized behaviors qualitatively resembling many biological phenomena. Leveraging the resulting archive, we then run a scaling study across six spatial scales and seven time horizons, uncovering macro-scale organization with no analogue at the base scale and characterizing how goal-space metrics behave at scale. This illustrates a strength of our approach: a relatively cheap large-scale diversity search can act as a principled scaffold for designing subsequent, more expensive experiments, enabling an iterative loop of experiment design, inspection, and redesign, supported by an interactive exploration tool that keeps scientists in the loop. Though demonstrated with Flow-Lenia, this approach potentially applies to other parameterizable complex systems where studying bottom-up collective behavior is of interest.
Abstract:We present a curiosity-driven AI scientist method for discovering system-level dynamics in Flow-Lenia, a continuous cellular automaton (CA) with mass conservation and parameter localization. Building on prior work that uses diversity search in Lenia to find individual self-organized patterns, we adapt Intrinsically Motivated Goal Exploration Processes (IMGEPs) to large environments of interacting patterns, using simulation-wide metrics such as evolutionary activity, compression ratio, and multi-scale matter distribution. We apply IMGEP in two exploration experiments: one targeting ecosystem-level dynamics, the other matter movement through obstacle-laden environments. In both, IMGEP illuminates significantly more of the metric space than random search and reveals self-organized behaviors qualitatively resembling many biological phenomena. Leveraging the resulting archive, we then run a scaling study across six spatial scales and seven time horizons, uncovering macro-scale organization with no analogue at the base scale and characterizing how goal-space metrics behave at scale. This illustrates a strength of our approach: a relatively cheap large-scale diversity search can act as a principled scaffold for designing subsequent, more expensive experiments, enabling an iterative loop of experiment design, inspection, and redesign, supported by an interactive exploration tool that keeps scientists in the loop. Though demonstrated with Flow-Lenia, this approach potentially applies to other parameterizable complex systems where studying bottom-up collective behavior is of interest.
Extended version of the paper first published at ALife 2025. Project webpage: this https URL 24 pages, 16 figures
Artificial Intelligence (cs.AI)
arXiv:2505.15998 [cs.AI]
(or arXiv:2505.15998v4 [cs.AI] for this version)
Proceedings of the Artificial Life Conference 2025, pp. 633-643
https://doi.org/10.1162/ISAL.a.896 DOI(s) linking to related resources
Text integral sub licența deschisă a publicației · 255 cuvinte