Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
To tackle the narrative collapse that typically occurs when scaling LLM-generated fiction beyond 10,000 words, this work evaluates story generation out to full-length novel scales of 100,000 words. The authors deploy NstAgent, a training-free agentic framework that dynamically maintains an explicit narrative state composed of character profiles, chronological events, and future plot obligations. Across systematic evaluations on extended consistency benchmarks, NstAgent prevents the expected quality degradation at scale, maintaining both writing quality and plot coherence from 10K to 100K words.
LLMs can scale to full 100,000-word novels without narrative collapse simply by offloading character arcs, plot history, and future constraints to an explicit, training-free state-tracking agent.
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, we introduce 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. We 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. We show that NstAgent achieves better narrative consistency and writing quality as stories grow longer, and neither of them degrades noticeably as length increases, suggesting that it provides an effective approach to scaling story generation toward full-length novels.
Tsinghua AI