arXiv cs.AI· Ga\"etan Berlaimont, Justin Vanwichelen, Beno\^it Ronval, H\'el\`ene Verhaeghe·· 2 天前AI 评分27
CriticaLLM：用 LLM 作为桌面角色扮演游戏的自主游戏主持人
CriticaLLM: Using LLMs as Autonomous Game Master for Tabletop Role-Playing Games
AI 导读
CriticaLLM 是一个全自动游戏主持人框架，可运行完整 TTRPG 游戏，从生成多样化故事到管理探索与战斗阶段。研究用真实 TTRPG 玩家进行评估，并考察四种不同规模 LLM 生成剧情与反派角色的多样性。
正文 · 原文
Abstract:Large Language Models (LLMs) have shown impressive capabilities across many tasks, including processing large amounts of text, following instructions, and generating ideas. This allows the use of such models in various contexts, including games and Tabletop Role-Playing Games (TTRPGs), such as Dungeons & Dragons. However, these games include many complex components, from crafting interesting stories to adhering to rules, while allowing some freedom in players' actions. In this work, we first highlight key components that LLMs should exhibit in the context of TTRPGs. Then, we introduce our framework, CriticaLLM, a fully automated Game Master that can run entire TTRPG games, from creating diverse and interesting stories to managing exploration and combat phases. We evaluate our framework with actual TTRPG players and examine the diversity of generated plots and antagonists across four LLMs of varying sizes. Overall, our framework appears to be a promising option for running entire TTRPG games.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.21525 [cs.AI] |
| (or arXiv:2610.21525v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.21525 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Benoît Ronval [view email]
[v1]
Thu, 15 Oct 2026 08:56:12 UTC (321 KB)
来源：arXiv cs.AI · arxiv.org