Paper Overview
Field: NLP Authors: Zhennan Wan, Jianfei Chen Published: 2026-09-28 arXiv: 2609.35759
Summary
LLMs have demonstrated strong capabilities in creative writing. However, scaling them to full-length novels remains challenging, as maintaining narrative consistency becomes increasingly difficult. Existing story-generation methods typically focus on stories of up to about ten thousand words, leaving their ability to scale to full-length novels underexplored.
In this work, the authors introduce the Narrative State Tracking Agent (NstAgent), a training-free agentic framework that allows LLMs to track a structured narrative state including characters, past events, and future requirements. They extend an existing benchmark to compare narrative consistency across lengths, and use it together with a writing-quality benchmark to systematically evaluate stories ranging from 10K to 100K words.
Key Findings
- NstAgent achieves better narrative consistency and writing quality as stories become longer.
- Neither consistency nor writing quality degrades noticeably as length increases from 10K to 100K words.
- The results suggest NstAgent provides an effective approach for scaling story generation to full-length novels.
- arXiv: https://arxiv.org/abs/2609.35759