Human Cognition in Machines: A Unified Perspective of World Models

TL;DR

This paper proposes a unified framework for world models in AI, integrating cognitive functions from Cognitive Architecture Theory, and highlights underexplored areas like motivation and meta-cognition.

Contribution

It introduces a comprehensive taxonomy and the concept of Epistemic World Models, guiding future research in cognitive functions and scientific discovery in AI.

Findings

Identifies motivation and meta-cognition as under-researched areas.

Proposes concrete directions based on active inference and global workspace theory.

Introduces Epistemic World Models for scientific discovery.

Abstract

This comprehensive report distinguishes prior works by the cognitive functions they innovate. Many works claim an almost "human-like" cognitive capability in their world models. To evaluate these claims requires a proper grounding in first principles in Cognitive Architecture Theory (CAT). We present a conceptual unified framework for world models that fully incorporates all the cognitive functions associated with CAT (i.e. memory, perception, language, reasoning, imagining, motivation, and meta-cognition) and identify gaps in the research as a guide for future states of the art. In particular, we find that motivation (especially intrinsic motivation) and meta-cognition remain drastically under-researched, and we propose concrete directions informed by active inference and global workspace theory to address them. We further introduce Epistemic World Models, a new category encompassing agent frameworks for scientific discovery that operate over structured knowledge. Our taxonomy, applied across video, embodied, and epistemic world models, suggests research directions where prior taxonomies have not.