Researchers have developed Shadow-Loom, an open-source framework designed to create graphical world models from narratives. This system employs causal physics based on Judea Pearl's work and a narrative physics engine to analyze reader engagement through states like mystery and suspense. While large language models are used for boundary tasks such as extraction and audit, the core reasoning processes are handled by typed code operating on the graph. AI
Summary written by gemini-2.5-flash-lite from 2 sources. How we write summaries →
IMPACT Introduces a novel framework for narrative analysis that separates core reasoning from LLM boundary tasks.
RANK_REASON The cluster describes an academic paper detailing an experimental open-source framework for narrative analysis.