When Should a World Model Move? Loss-Conditioned State Execution
作者: Jintao Xu, Zhengyu Chen, Ben Zhang, Yongzhi Qi, Jianshen Zhang
分类: cs.AI, cs.LG, math.OC
发布日期: 2026-09-14
💡 一句话要点
提出损失条件状态执行以优化世界模型的状态更新
🎯 匹配领域: 支柱二：RL算法与架构 (RL & Architecture)
关键词: 状态执行 损失条件 世界模型 动态决策 预测优化 机器学习 智能系统
📋 核心要点
- 现有方法在执行世界模型提案时，无法有效判断更新是否会降低损失，导致决策不够准确。
- 论文提出了一种损失条件状态执行的方法，通过构建损失特定的可行提案来优化状态更新决策。
- 在28,684个M4月度系列的实验中，该方法实现了0.588的有界损失，相比于持久性方法的0.599有显著提升。
📝 摘要（中文）
本文介绍了一种损失条件状态执行的方法，该方法能够决定是否执行世界模型的固定可行提案或保留当前状态。仅凭预测信息无法确定更新是否会降低下游损失。我们将状态可移动性形式化为存在一个减少损失的可行修正，并将其与特定提案的收益区分开来。实验结果表明，该方法在多个基准测试中表现出色，能够有效平衡认证和覆盖。
🔬 方法详解
问题定义：本文旨在解决如何有效判断世界模型的状态更新是否能降低损失的问题。现有方法往往依赖于预测信息，未能充分考虑更新的实际效果。
核心思路：提出损失条件状态执行的核心思路是构建一个损失特定的可行提案，并通过评估其在独立校准单元上的有界损失增益来决定是否执行。
技术框架：整体方法包括以下几个主要模块：首先，利用预测分布构建损失特定的可行提案；其次，评估提案在不同组别中的有界损失增益；最后，仅在具有正的同时下置信界的组别中执行提案。
关键创新：最重要的技术创新在于将状态可移动性形式化为存在一个减少损失的可行修正，并与特定提案的收益区分开来。这一创新使得决策过程更加科学和有效。
关键设计：在参数设置上，方法使用了独立同分布的校准单元，并通过统计方法确保每个接受的组别在高概率下具有低于持久性的期望损失。
🖼️ 关键图片
📊 实验亮点
在28,684个M4月度系列的实验中，该方法仅在14.0%的系列中执行提案，取得了0.588的有界损失，相比于持久性方法的0.599和始终执行提案的0.621，显示出显著的性能提升。95%的配对自助法区间均低于零，表明方法的有效性。
🎯 应用场景
该研究具有广泛的应用潜力，尤其在需要动态决策的领域，如金融预测、库存管理和智能机器人等。通过优化状态更新策略，可以显著提高系统的决策效率和准确性，推动相关领域的智能化发展。
📄 摘要（原文）
We introduce loss-conditioned state execution, a model-agnostic method that decides whether to execute a world model's fixed feasible proposal or retain the current state. Predictive informativeness alone, however, does not establish whether an update will reduce downstream loss. Occurrence ranking can approach perfection while persistence remains the unique absolute-loss Bayes action. Two transition laws can also share occurrence information and conditional variance yet require opposite absolute-loss decisions. We formalize state movability as the existence of a loss-reducing feasible correction and distinguish it from the benefit of a particular proposal. Our method constructs a loss-specific feasible proposal from a predictive distribution and evaluates its groupwise bounded-loss gain over persistence on independent calibration units. The proposal is executed only in groups with a positive simultaneous lower confidence bound. For fixed proposals and groups with bounded unit losses, we prove that every accepted group has lower expected loss than persistence with high probability when calibration units are i.i.d. draws from the target population. Experiments on public forecasting and action-conditioned dynamics benchmarks show supported updates and a trade-off between certification and coverage. On 28,684 held-out M4 Monthly series, the method executes the proposal for 14.0% of series and achieves bounded loss 0.588, compared with 0.599 for persistence and 0.621 for always executing the proposal. The paired 95% bootstrap intervals for both comparisons lie below zero. In constrained forecasting of six unhealthy-inventory types from JD$\mbox{.}$com, a leading e-retailer in China, strong occurrence-ranking signal coexists with a loss-based preference for persistence, illustrating why event predictability and state execution must be evaluated separately.