Multiverse of Greatness: Generating Story Branches with LLMs

TL;DR

This paper introduces Dynamic Context Prompting/Programming (DCP/P), a framework for generating coherent, graph-based story branches with LLMs by maintaining a dynamic context window, improving over static prompt methods.

Contribution

The paper proposes DCP/P, a novel framework that enhances story generation with LLMs by dynamically managing context, enabling more coherent and flexible story branching.

Findings

Providing proper context improves story quality.

LLMs exhibit word bias despite different models.

Dynamic context management outperforms static prompts.

Abstract

This paper presents Dynamic Context Prompting/Programming (DCP/P), a novel framework for interacting with LLMs to generate graph-based content with a dynamic context window history. While there is an existing study utilizing LLMs to generate a visual novel game, the previous study involved a manual process of output extraction and did not provide flexibility in generating a longer, coherent story. We evaluate DCP/P against our baseline, which does not provide context history to an LLM and only relies on the initial story data. Through objective evaluation, we show that simply providing the LLM with a summary leads to a subpar story compared to additionally providing the LLM with the proper context of the story. We also provide an extensive qualitative analysis and discussion. We qualitatively examine the quality of the objectively best-performing generated game from each approach. In addition, we examine biases in word choices and word sentiment of the generated content. We find a consistent observation with previous studies that LLMs are biased towards certain words, even with a different LLM family. Finally, we provide a comprehensive discussion on opportunities for future studies.