外观
论文arXiv 2610.0498824 天前
多久，而非多近：用于潜在世界模型规划的学习型时间度量
原文链接：How Long, Not How Close: A Learned Temporal Metric for Planning in Latent World Models（arXiv 2610.04988 · 发布日期：2026-10-04）
How Long, Not How Close: A Learned Temporal Metric for Planning in Latent World Models
针对远距离目标下潜在距离排序失效，TEMPO学习时间距离代价，并在十一种仿真环境与LeWM、PLDM上验证。
核心要点
- 潜在世界模型按想象终态与目标的潜在距离排序，目标相隔多个计划时会失效。
- TEMPO不改冻结世界模型，从已有轨迹学习反映环境步数的潜在映射并融入规划代价。
- 该方法无需奖励、策略或成功标签，且作为代价项对规划搜索的额外开销很小。
- 在十一种仿真环境中，TEMPO提升LeWM和PLDM，并在各目标距离下均有改进。
验证范围与局限：验证范围为十一种仿真环境，包括迷宫导航、桌面推动、机械臂控制和三维操作。
英文摘要 (Abstract)
Latent world models plan by rolling a frozen predictor forward under candidate action sequences and ranking the candidates by the latent distance between their imagined end state and the goal. However, this ranking breaks down when the goal lies several plans away, because the latent distance measures how closely an end state resembles the goal rather than how far it remains from reaching it. To address this, we propose TEMPO, a temporal-distance planning objective that leaves the world model untouched, learns only from the recorded trajectories already used to train it, and adds negligible cost to the planner's search. TEMPO learns a small map of the frozen latent in which the distance between two states of an episode reflects the number of environment steps between them, and blends this distance into the planner's cost. It requires no rewards, policies or success labels and, being a cost rather than a model, applies to frozen world models with one latent vector per state that plan by a latent distance. We evaluate TEMPO on eleven simulated environments (e.g., maze navigation, tabletop pushing, robotic arm control and three-dimensional manipulation) with the LeWM and PLDM planners. With a small MLP that adds at most 0.3% to a plan's arithmetic, TEMPO improves both planners at every goal distance, including the one-plan setting of their evaluations, raises LeWM from 36% to 99% on TwoRoom three plans from the goal, and remains competitive on a broad range of 2D and 3D navigation, reaching and manipulation tasks.
作者：Lama Moukheiber, Haotian Xue, Yongxin Chen
来源
本页事实对应的公开页面，以原文为准。