stable-worldmodel-v1: Reproducible World Modeling Research and Evaluation

TL;DR

This paper introduces stable-worldmodel (SWM), a comprehensive, standardized ecosystem for research on world models, facilitating reproducibility, robustness, and continual learning in environment modeling.

Contribution

SWM provides a modular, tested, and documented platform with tools, environments, and baselines to advance world model research and evaluation.

Findings

SWM enables standardized benchmarking of world models.

Demonstrated zero-shot robustness in DINO-WM using SWM.

Supports controllable environment variations for robustness studies.

Abstract

World Models have emerged as a powerful paradigm for learning compact, predictive representations of environment dynamics, enabling agents to reason, plan, and generalize beyond direct experience. Despite recent interest in World Models, most available implementations remain publication-specific, severely limiting their reusability, increasing the risk of bugs, and reducing evaluation standardization. To mitigate these issues, we introduce stable-worldmodel (SWM), a modular, tested, and documented world-model research ecosystem that provides efficient data-collection tools, standardized environments, planning algorithms, and baseline implementations. In addition, each environment in SWM enables controllable factors of variation, including visual and physical properties, to support robustness and continual learning research. Finally, we demonstrate the utility of SWM by using it to study zero-shot robustness in DINO-WM.