理解潜在世界模型的能力边界
Understanding the limits of latent world models
AI 导读
一项 arXiv 研究提出“忠实潜在世界模型”概念，将其定义为环境转移映射与学到的潜在动力学之间的等变性条件，并提出潜在滚动一致性（LRC）指标来评估动作条件下的潜在预测。研究发现，所研究的当前最先进潜在世界模型中，没有一个是通过忠实建模环境动力学来真正解决任务的。
正文 · 原文
Abstract:World models are widely proposed as core components for enabling agents to reason over long temporal and spatial horizons. Yet current evaluations often emphasize task performance while providing limited insight into whether learned world models acquire predictive latent dynamics that reflect the underlying structure of the environment. Here we introduce the concept of faithful latent world models as an equivariance condition between the environment's transition map and the learned latent dynamics. We propose latent rollout consistency (LRC) as a metric to evaluate action-conditioned latent prediction. Strikingly, we find that across the studied state of the art latent world models, none solves the task by actually faithfully modeling the dynamics of the environment.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.37588 [cs.LG] |
| (or arXiv:2610.37588v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.37588 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Rodrigo Gonzalez Laiz [view email]
[v1]
Tue, 27 Oct 2026 17:25:50 UTC (17,985 KB)
来源：arXiv cs.AI · arxiv.org