多少规划才算足够？在世界模型规划中减少搜索与计算

中文摘要 AI 辅助

视觉世界模型通过决策时动作搜索实现目标导向控制，但其部署效率常受制于保守的较大规划预算。我们证明，无需与全预算动作保持一致即可获得具有竞争力的任务性能，足够的预算在不同模型-任务对之间有所差异，且迭代规划器会反复编码与求解无关的上下文。为解决这些低效问题，我们提出SufficientPlan，一个简单的部署框架，无需修改预训练世界模型或规划器更新。其配对顺序预算认证（PSBC）组件使用配对闭环证据，在预定义的全性能容差内搜索并认证一个缩减的模型-任务特定预算。其静态上下文重用（SCR）组件在搜索迭代间缓存观测和目标表示，同时保留依赖候选的规划和所选动作。在多个世界模型骨干和视觉控制任务上的实验表明，SufficientPlan在保持竞争力控制性能的同时，大幅减少了搜索预算和规划延迟。

英文摘要

Visual world models enable goal-directed control through decision-time action search, but their deployment efficiency is often limited by conservatively large planning budgets. We show that competitive task performance can be achieved without agreement with the Full-budget action, that sufficient budgets vary across model--task pairs, and that iterative planners repeatedly encode solve-invariant context. To address these inefficiencies, we propose {SufficientPlan}, a simple deployment framework that requires no modification to pretrained world models or planner updates. Its {Paired Sequential Budget Certification (PSBC)} component uses paired closed-loop evidence to search for and certify a reduced model--task-specific budget within a predefined Full-performance tolerance. Its {Static-Context Reuse (SCR)} component caches observation and goal representations across search iterations while preserving candidate-dependent planning and selected actions. Experiments across multiple world-model backbones and visual-control tasks show that SufficientPlan substantially reduces search budgets and planning latency while maintaining competitive control performance.

发表机构

• Beihang University（北京航空航天大学）
• Nanyang Technological University（南洋理工大学）
• Chengdu University of Technology（成都理工大学）

机构由 AI 辅助整理，请以论文原文为准。