Word2World: Generating Stories and Worlds through Large Language Models

TL;DR

Word2World leverages large language models to procedurally generate playable game worlds and stories without task-specific fine-tuning, demonstrating the potential of LLMs in game content creation.

Contribution

This work introduces Word2World, a novel system that enables LLMs to design game worlds and narratives procedurally without fine-tuning, combining content creation and information extraction capabilities.

Findings

Effective generation of game stories and worlds by LLMs.

Validation through ablation studies confirms each component's contribution.

Open-source implementation available for further research.

Abstract

Large Language Models (LLMs) have proven their worth across a diverse spectrum of disciplines. LLMs have shown great potential in Procedural Content Generation (PCG) as well, but directly generating a level through a pre-trained LLM is still challenging. This work introduces Word2World, a system that enables LLMs to procedurally design playable games through stories, without any task-specific fine-tuning. Word2World leverages the abilities of LLMs to create diverse content and extract information. Combining these abilities, LLMs can create a story for the game, design narrative, and place tiles in appropriate places to create coherent worlds and playable games. We test Word2World with different LLMs and perform a thorough ablation study to validate each step. We open-source the code at https://github.com/umair-nasir14/Word2World.