Demo Paper: A Game Agents Battle Driven by Free-Form Text Commands Using Code-Generation LLM
Ray Ito, Junichiro Takahashi
TL;DR
The paper addresses the limitation of rule-based command systems in games by enabling flexible, language-driven control of game agents through translation of natural-language commands into behavior branches via a code-generation LLM. It presents an end-to-end system that integrates Unity, real-time networking, and an AWS backend (Cognito, Lambda, DynamoDB) to generate and execute these behavior branches using the llama-v2-34b-code model via the Fireworks AI API, with logs stored for validation. The contribution includes a complete system architecture, latency measurements, and a live two-player demonstration, illustrating practical feasibility and offering a data pipeline for future improvements. This work has practical significance for the game industry by enabling more natural and responsive agent control and providing a foundation for data-driven refinement of language-driven game AI.
Abstract
This paper presents a demonstration of our monster battle game, in which the game agents fight in accordance with their player's language commands. The commands were translated into the knowledge expression called behavior branches by a code-generation large language model. This work facilitated the design of the commanding system more easily, enabling the game agent to comprehend more various and continuous commands than rule-based methods. The results of the commanding and translation process were stored in a database on an Amazon Web Services server for more comprehensive validation. This implementation would provide a sufficient evaluation of this ongoing work, and give insights to the industry that they could use this to develop their interactive game agents.