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

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Maes, Lucas, Lidec, Quentin Le, Haramati, Dan, Massaudi, Nassim, Scieur, Damien, LeCun, Yann, Balestriero, Randall
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912910346813440
author Maes, Lucas
Lidec, Quentin Le
Haramati, Dan
Massaudi, Nassim
Scieur, Damien
LeCun, Yann
Balestriero, Randall
author_facet Maes, Lucas
Lidec, Quentin Le
Haramati, Dan
Massaudi, Nassim
Scieur, Damien
LeCun, Yann
Balestriero, Randall
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2602_08968
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle stable-worldmodel-v1: Reproducible World Modeling Research and Evaluation
Maes, Lucas
Lidec, Quentin Le
Haramati, Dan
Massaudi, Nassim
Scieur, Damien
LeCun, Yann
Balestriero, Randall
Artificial Intelligence
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.
title stable-worldmodel-v1: Reproducible World Modeling Research and Evaluation
topic Artificial Intelligence
url https://arxiv.org/abs/2602.08968