الملخص
Interactive world models require responding in real time to versatile controls and maintaining long-horizon consistency. However, modeling heterogeneous controls remains difficult, while explosive contexts and unstable distillation impede achieving both long-horizon consistency and real-time responsiveness. In this paper, we present WorldPlay2, an interactive world model that couples a factorized hybrid control interface with a co-design of compressed memory and stable distillation. 1) Our factorized hybrid control interface integrates frame-aligned action control with structured semantic control that explicitly disentangles scene appearance, character identity, and dynamic semantic events, thereby facilitating effective control learning. 2) To achieve efficient long-horizon modeling, we compress historical contexts into compact memory tokens shared by the autoregressive student and the bidirectional teacher. This design enables clip-wise, memory-conditioned score evaluation instead of jointly processing an entire long rollout, substantially reducing distillation overhead. 3) We further propose Stable Forcing, which initializes the autoregressive student via a few-step strategy and leverages full-rollout replay to preserve the quality of long-horizon rollouts, ensuring robust and stable distillation. Extensive experiments demonstrate the strong generalizability of our model and its superior performance compared to existing methods.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Zhang, H., Sun, W., Wang, T., Wu, J., Zhang, J., Wang, Y., Qiao, Y., & Guo, C. (2026). WorldPlay2: Extending Real-Time Interactive World Models in Control and Horizon. https://omanscience.com/ar/articles/worldplay2-extending-real-time-interactive-world-models-in-control-and-horizon
MLA 9
Zhang, Haiyu, et al. "WorldPlay2: Extending Real-Time Interactive World Models in Control and Horizon." https://omanscience.com/ar/articles/worldplay2-extending-real-time-interactive-world-models-in-control-and-horizon.
شيكاغو (المؤلف–التاريخ)
Zhang, Haiyu, Wenqiang Sun, Tengfei Wang, Junta Wu, Jun Zhang, Yunhong Wang, Yu Qiao, and Chunchao Guo. 2026. "WorldPlay2: Extending Real-Time Interactive World Models in Control and Horizon." https://omanscience.com/ar/articles/worldplay2-extending-real-time-interactive-world-models-in-control-and-horizon.
هارفارد
Zhang, H., Sun, W., Wang, T., Wu, J., Zhang, J., Wang, Y., Qiao, Y. and Guo, C. (2026) 'WorldPlay2: Extending Real-Time Interactive World Models in Control and Horizon', Available at: https://omanscience.com/ar/articles/worldplay2-extending-real-time-interactive-world-models-in-control-and-horizon.
فانكوفر
Zhang H, Sun W, Wang T, Wu J, Zhang J, Wang Y, et al. WorldPlay2: Extending Real-Time Interactive World Models in Control and Horizon. https://omanscience.com/ar/articles/worldplay2-extending-real-time-interactive-world-models-in-control-and-horizon
IEEE
H. Zhang, W. Sun, T. Wang, J. Wu, J. Zhang, Y. Wang, Y. Qiao, and C. Guo, "WorldPlay2: Extending Real-Time Interactive World Models in Control and Horizon," https://omanscience.com/ar/articles/worldplay2-extending-real-time-interactive-world-models-in-control-and-horizon.