并行预测世界模型用于准确且高效的长时域规划
Parallel Predictive World Models for Accurate and Efficient Long-Horizon Planning
- Tsinghua University（清华大学）
- Institute of Microelectronics, Chinese Academy of Sciences（中国科学院微电子研究所）
- The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））
- Nanyang Technological University, Singapore（新加坡南洋理工大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本文提出并行预测世界模型（PPWM），通过并行预测轨迹并保留因果交互，避免了自回归的递归反馈，在视觉控制任务中实现更低预测误差、更高规划成功率及超过3倍的规划加速。
AI中文摘要：
长时域世界模型规划通常依赖于自回归展开，其中预测的状态被反复反馈到模型中。这保持了时间结构，但产生了一条长度等于时域长度的顺序路径，并使后续预测暴露于递归的解码状态反馈中。我们引入了并行预测世界模型（PPWM），它在并行预测有限时域轨迹的同时，保留未来表示之间的因果交互。每个时域步骤以其因果动作前缀为条件，未来表示在解码前进行交互，从而将时间因果性与逐状态输出递归分离。我们通过将自回归展开视为因果轨迹映射，并识别出PPWM所消除的解码状态反馈路径，来形式化这一区别。在四个视觉控制任务中，PPWM在评估的预测接口中实现了最低的长时域预测误差和最高的交叉熵方法（CEM）模拟器成功率。同时，PPWM相较于自回归的LeWM基线，实现了超过3倍的平均CEM规划加速。这些结果表明，准确且高效的长时域世界模型规划并不需要逐状态自回归，而是可以通过并行因果轨迹预测来实现。
英文摘要：
Long-horizon world-model planning typically relies on autoregressive rollouts, where predicted states are repeatedly fed back into the model. This preserves temporal structure but creates a horizon-length sequential path and exposes later predictions to recursive decoded-state feedback. We introduce Parallel Predictive World Models (PPWM), which predict a finite-horizon trajectory in parallel while retaining causal interaction among future representations. Each horizon is conditioned on its causal action prefix, and future representations interact before decoding, separating temporal causality from state-by-state output recursion. We formalize this distinction by viewing autoregressive rollout as a causal trajectory map and identifying the decoded-state feedback pathway removed by PPWM. Across four visual-control tasks, PPWM achieves the lowest long-horizon prediction error and the highest Cross-Entropy Method (CEM) simulator success among the evaluated predictive interfaces. Meanwhile, PPWM achieves more than a 3$\times$ average CEM planning speedup over the autoregressive LeWM baseline. These results suggest that accurate and efficient long-horizon world-model planning does not require state-by-state autoregression, but can instead be achieved through parallel causal trajectory prediction.
