Level Generation Through Large Language Models

Abstract

Large Language Models (LLMs) are powerful tools, capable of leveraging their training on natural language to write stories, generate code, and answer questions. But can they generate functional video game levels? Game levels, with their complex functional constraints and spatial relationships in more than one dimension, are very different from the kinds of data an LLM typically sees during training. Datasets of game levels are also hard to come by, potentially taxing the abilities of these data-hungry models. We investigate the use of LLMs to generate levels for the game Sokoban, finding that LLMs are indeed capable of doing so, and that their performance scales dramatically with dataset size. We also perform preliminary experiments on controlling LLM level generators and discuss promising areas for future work.

References in corpus (2)

• Deep Learning for Procedural Content Generation
• Super Mario as a String: Platformer Level Generation Via LSTMs

Cited by in corpus (8)

• LLMatic: Neural Architecture Search via Large Language Models and Quality Diversity Optimization
• GPT for Games: A Scoping Review (2020-2023)
• Generative Design through Quality-Diversity Data Synthesis and Language Models
• Lottery and Sprint: Generate a Board Game with Design Sprint Method on AutoGPT
• FlowGPT: Exploring Domains, Output Modalities, and Goals of Community-Generated AI Chatbots
• Not All the Same: Understanding and Informing Similarity Estimation in Tile-Based Video Games
• Boardwalk: Towards a Framework for Creating Board Games with LLMs
• Analysis of Robustness of a Large Game Corpus