Building spatial world models from sparse transitional episodic memories

TL;DR

This paper introduces ESWM, a novel model that constructs flexible spatial maps from sparse episodic memories, enabling rapid adaptation and efficient navigation without extensive training.

Contribution

The paper presents ESWM, a new framework that builds spatial world models from disjoint episodic memories, inspired by neuroscience, allowing quick environmental adaptation and planning.

Findings

ESWM predicts unobserved transitions from minimal experience.

The latent space geometry aligns with environment structure.

Enables near-optimal exploration and navigation strategies.

Abstract

Many animals possess a remarkable capacity to rapidly construct flexible cognitive maps of their environments. These maps are crucial for ethologically relevant behaviors such as navigation, exploration, and planning. Existing computational models typically require long sequential trajectories to build accurate maps, but neuroscience evidence suggests maps can also arise from integrating disjoint experiences governed by consistent spatial rules. We introduce the Episodic Spatial World Model (ESWM), a novel framework that constructs spatial maps from sparse, disjoint episodic memories. Across environments of varying complexity, ESWM predicts unobserved transitions from minimal experience, and the geometry of its latent space aligns with that of the environment. Because it operates on episodic memories that can be independently stored and updated, ESWM is inherently adaptive, enabling rapid adjustment to environmental changes. Furthermore, we demonstrate that ESWM readily enables near-optimal strategies for exploring novel environments and navigating between arbitrary points, all without the need for additional training. Our work demonstrates how neuroscience-inspired principles of episodic memory can advance the development of more flexible and generalizable world models.