Abstract
Embodied world models enable scalable simulation of embodied interactions for robot learning. However, existing models are prone to Model Laziness, as they focus on visual fidelity at the expense of physical reasoning and lack process-level supervision over the temporal dynamics of manipulated objects. In this work, we propose EVEWorld, a physical evolution-supervision framework for physically consistent target evolution. EVEWorld consists of two components: Instance-Guided Restoration (IGR) and Temporal Instance Alignment (TIA). First, IGR promotes instance consistency through restoration supervision. Second, TIA promotes cross-frame consistency by aligning target instances across adjacent frames. We further introduce the Model Laziness Rate (MLR), a metric that measures persistent violations of instance consistency in generated trajectories. Extensive experiments on DreamGenBench, EWMBench, and PBench demonstrate the effectiveness of EVEWorld, notably achieving an 87.5% reduction in MLR compared with GigaWorld-0. On the WorldArena 2.0 Track 1 leaderboard, our model ranks 6th in JEPA Similarity and 17th overall, which further validates the performance of our evolution supervision strategy.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Wang, K., Zhang, S., Xie, Z., Xiong, X., Song, Z., Wu, Z., Lu, J., Teng, Y., Song, Z., & Zhang, J. (2026). EVEWorld: Physical Evolution Supervision for Embodied World Models. https://omanscience.com/en/articles/eveworld-physical-evolution-supervision-for-embodied-world-models
MLA 9
Wang, Kaiqi, et al. "EVEWorld: Physical Evolution Supervision for Embodied World Models." https://omanscience.com/en/articles/eveworld-physical-evolution-supervision-for-embodied-world-models.
Chicago (author–date)
Wang, Kaiqi, Songxin Zhang, Zejian Xie, Xiao Xiong, Zhuoyang Song, Ziwei Wu, Junyu Lu, Yitan Teng, Ziying Song, and Jiaxing Zhang. 2026. "EVEWorld: Physical Evolution Supervision for Embodied World Models." https://omanscience.com/en/articles/eveworld-physical-evolution-supervision-for-embodied-world-models.
Harvard
Wang, K., Zhang, S., Xie, Z., Xiong, X., Song, Z., Wu, Z., Lu, J., Teng, Y., Song, Z. and Zhang, J. (2026) 'EVEWorld: Physical Evolution Supervision for Embodied World Models', Available at: https://omanscience.com/en/articles/eveworld-physical-evolution-supervision-for-embodied-world-models.
Vancouver
Wang K, Zhang S, Xie Z, Xiong X, Song Z, Wu Z, et al. EVEWorld: Physical Evolution Supervision for Embodied World Models. https://omanscience.com/en/articles/eveworld-physical-evolution-supervision-for-embodied-world-models
IEEE
K. Wang, S. Zhang, Z. Xie, X. Xiong, Z. Song, Z. Wu, J. Lu, Y. Teng, Z. Song, and J. Zhang, "EVEWorld: Physical Evolution Supervision for Embodied World Models," https://omanscience.com/en/articles/eveworld-physical-evolution-supervision-for-embodied-world-models.