Hierarchically Organized Latent Modules for Exploratory Search in Morphogenetic Systems
Mayalen Etcheverry,Clu00e9ment Moulin-Frier,Pierre-Yves Oudeyer
Self-organization of complex morphological patterns from local interactions is afascinating phenomenon in many natural and artificial systems. In the artificialworld, typical examples of such morphogenetic systems are cellular automata. Yet,their mechanisms are often very hard to grasp and so far scientific discoveries ofnovel patterns have primarily been relying on manual tuning and ad hoc exploratorysearch. The problem of automated diversity-driven discovery in these systems wasrecently introduced [26, 62], highlighting that two key ingredients are autonomousexploration and unsupervised representation learning to describe u201crelevantu201d degreesof variations in the patterns. In this paper, we motivate the need for what we callMeta-diversity search, arguing that there is not a unique ground truth interestingdiversity as it strongly depends on the final observer and its motives. Using acontinuous game-of-life system for experiments, we provide empirical evidencesthat relying on monolithic architectures for the behavioral embedding design tendsto bias the final discoveries (both for hand-defined and unsupervisedly-learnedfeatures) which are unlikely to be aligned with the interest of a final end-user. Toaddress these issues, we introduce a novel dynamic and modular architecture thatenables unsupervised learning of a hierarchy of diverse representations. Combinedwith intrinsically motivated goal exploration algorithms, we show that this systemforms a discovery assistant that can efficiently adapt its diversity search towardspreferences of a user using only a very small amount of user feedback.
