Language Models, Agent Models, and World Models: The LAW for Machine Reasoning and Planning

TL;DR

This paper introduces the LAW framework, integrating language, agent, and world models to enhance machine reasoning and planning, addressing limitations of language models in complex scenarios.

Contribution

It proposes a novel abstraction for reasoning that combines world and agent models with language models as a backend, offering a new perspective for robust AI reasoning.

Findings

World and agent models improve reasoning capabilities.

Language models serve as adaptable computational backends.

Review of recent progress and future research directions.

Abstract

Despite their tremendous success in many applications, large language models often fall short of consistent reasoning and planning in various (language, embodied, and social) scenarios, due to inherent limitations in their inference, learning, and modeling capabilities. In this position paper, we present a new perspective of machine reasoning, LAW, that connects the concepts of Language models, Agent models, and World models, for more robust and versatile reasoning capabilities. In particular, we propose that world and agent models are a better abstraction of reasoning, that introduces the crucial elements of deliberate human-like reasoning, including beliefs about the world and other agents, anticipation of consequences, goals/rewards, and strategic planning. Crucially, language models in LAW serve as a backend to implement the system or its elements and hence provide the computational power and adaptability. We review the recent studies that have made relevant progress and discuss future research directions towards operationalizing the LAW framework.