Platonic Representation Hypothesis on World Models

世界模型中的柏拉图表征假设

Abstract: World models have demonstrated significant potential for perceiving and simulating complex environments. Despite their strong performance, the fundamental nature of their learned representations remains poorly understood.

摘要： 世界模型在感知和模拟复杂环境方面展现出了巨大的潜力。尽管它们表现出色，但其所学表征的本质仍未被充分理解。

In this paper, we investigate the Platonic Representation Hypothesis within this domain by proposing the Predictive Consistency Assumption: we posit that the optimization of a shared state transition objective acts as a selective pressure that encourages heterogeneous models to converge toward a shared latent structure.

在本文中，我们通过提出“预测一致性假设”（Predictive Consistency Assumption）来探讨该领域中的“柏拉图表征假设”：我们认为，对共享状态转移目标的优化起到了一种选择压力，促使异构模型向共享的潜在结构收敛。

Through systematic experiments with the DINO World Model (DINO-WM), in which we vary visual encoders to create heterogeneous models, we find that capable world models evolve toward geometrically similar internal structures.

通过对 DINO 世界模型（DINO-WM）进行系统性实验，我们改变视觉编码器以创建异构模型，结果发现，高性能的世界模型会向几何结构相似的内部结构演化。

Moreover, via model stitching, we show that the internal features of one world model can be mapped to another with limited performance degradation, providing evidence of functional compatibility.

此外，通过模型拼接（model stitching）实验，我们证明了一个世界模型的内部特征可以映射到另一个模型，且性能下降有限，这为功能兼容性提供了证据。

Our findings suggest that the pursuit of predictive consistency can promote shared, transition-compatible latent structure across world models.

我们的研究结果表明，对预测一致性的追求可以促进不同世界模型之间形成共享的、具有转移兼容性的潜在结构。