World-State Transformations for Neuro-symbolic Interactive Storytelling
Summary
This paper explores using LLMs to predict state changes within rule-based interactive storytelling systems, aiming to improve coherence and player expression. Experiments with Llama 3 70B and Gemini 1.5 Flash show that world-state transformations can maintain consistency while encouraging creative player input.
View Cached Full Text
Cached at: 05/26/26, 09:06 AM
# World-State Transformations for Neuro-symbolic Interactive Storytelling Source: [https://arxiv.org/abs/2605.24719](https://arxiv.org/abs/2605.24719) [View PDF](https://arxiv.org/pdf/2605.24719) > Abstract:Large Language Models \(LLMs\) have changed the possibilities of Interactive Storytelling systems that process free\-text user input\. However, as more of these systems are built, evidence continues to mount regarding the story coherence problems that arise when relying solely on them\. Recent research suggests that LLMs can effectively predict state changes within rule\-based Interactive Storytelling systems, triggering pre\-programmed world\-state transformations\. In this paper, we conduct an exploratory evaluation of whether such transformations can serve as a catalyst for player expression while aiming to address the incoherence issues typical of purely LLM\-based approaches\. Building upon a neuro\-symbolic architecture, we conducted experiments using an open\-source model \(Llama 3 70B\) and a closed\-source model \(Gemini 1\.5 Flash\), with testing conducted in both English and Spanish\. Eight participants played two scenarios, carefully designed to assess different evaluation objectives\. Our observations suggest that transformations offer a way to maintain world\-state consistency while encouraging players to interact creatively through their written inputs\. ## Submission history From: Santiago Góngora \[[view email](https://arxiv.org/show-email/3b65268f/2605.24719)\] **\[v1\]**Sat, 23 May 2026 20:14:39 UTC \(2,758 KB\)
Similar Articles
When Stories Evolve: Benchmarking LLM Storytelling Across Agent Architectures in Open-Ended World Simulations
This paper introduces WSE-bench, a process benchmark for evaluating LLMs in open-ended world simulations, separately assessing sustained generation, canonical coherence, and meaningful development.
ConWriter: Transition-Constrained Stateful Long-Form Story Generation with Lightweight Neuro-Symbolic Consistency Control
ConWriter introduces a training-free framework for long-form story generation that maintains narrative consistency through scene-level incremental writing, symbolic state reasoning, and uncertainty-aware risk signals. Evaluated on ConStory-Bench across multiple models and lengths, it aims to prevent consistency errors from propagating in extended contexts.
EvolvingWorld: An Open-Schema Framework for Co-Evolving Role-Play Agents and World Model in Interactive Literary World
Introduces EvolvingWorld, an open-schema framework and benchmark for co-evolving role-play agents and world models in interactive literary worlds, enabling long-horizon simulation with persistent character and world state updates.
PatchBoard: Schema-Grounded State Mutation for Reliable and Auditable LLM Multi-Agent Collaboration
PatchBoard replaces natural-language dialogue in LLM multi-agent systems with validated JSON Patch mutations over a shared structured state, achieving higher success rates and significantly lower token usage on ALFWorld benchmarks.
Can LLM Agents Stick to the Script? A Benchmark for Long-Horizon Consistency in Interactive Narratives
This paper introduces NCP-Bench, a benchmark derived from 100 movie synopses for evaluating long-horizon narrative consistency in LLM-based interactive storytelling agents. Experiments show that even strong models like GPT-5.2 struggle to maintain logical consistency, with a 42% survival rate after 20 turns and high fact conflict rates.