论文提出 LRC 指标：现有 latent world models 均未忠实建模环境动态
Understanding the limits of latent world models
研究者用新指标 LRC 检验了一圈主流 latent world models，结论挺扎眼：没一个真在学环境动态。做 world model 方向的值得看下。
这篇 arXiv 论文提出“忠实 latent world models”概念，定义为环境转移映射与学习到的 latent 动态之间的等变性条件。作者设计了 latent rollout consistency（LRC）指标，用于评估动作条件下的 latent 预测一致性。实验覆盖了所研究的多个 state of the art latent world models，结果显示没有一个模型是通过忠实建模环境动态来完成任务。该工作指出了当前以任务表现为主的评测方式掩盖了 latent 动态与真实环境结构的偏差。
Understanding the limits of latent world models
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.
