GENEVA: GENErating and Visualizing Branching Narratives Using LLMs

Dialogue-based Role Playing Games (RPGs) require powerful storytelling. Thenarratives of these may take years to write and typically involve a largecreative team. In this work, we demonstrate the potential of large generativetext models to assist this process. GENEVA, a prototype tool,generates a rich narrative graph with branching and reconverging storylinesthat match a high-level narrative description and constraints provided by thedesigner. A large language model (LLM), GPT-4, is used to generate thebranching narrative and to render it in a graph format in a two-step process.We illustrate the use of GENEVA in generating new branching narratives for fourwell-known stories under different contextual constraints. This tool has thepotential to assist in game development, simulations, and other applicationswith game-like properties.