Narrative World Model: Narratology-Grounded Writer Memory for Long-Form Fiction

Long-form fiction writers need memory that answers multi-hop questions about evolving story state: who knows a secret and when they learned it, whether an event preceded the narration that revealed it, whether a setup paid off, and how a relationship shifted. General-purpose retrieval and agent-memory systems represent entities and facts but do not necessarily make the narratological structure these questions turn on first-class, so they can surface the wrong evidence or none at all. We introduce the Narrative World Model (NWM), a writer-memory system that pairs a narratology-grounded typed temporal-state graph with query-conditioned hybrid retrieval. To measure memory rather than the answerer, we read every system through a single held-constant Opus~4.8 reader over only that system's chapter-safe evidence, on a reproducible public corpus and a model-adjudicated private multi-hop benchmark, and we compare against a strong temporal-knowledge-graph agent-memory framework, Graphiti/Zep. NWM substantially outperforms this baseline on the private multi-hop set and the full public set; the smaller public multi-hop comparison has the same direction but is not statistically significant. The gap survives rebuilding the baseline with NWM's own extractor and is consistent with fine-grained narrative decomposition and query-conditioned retrieval rather than graph size or extractor identity.