Researchers have developed EvolvingWorld, a novel framework designed for the co-evolution of characters and world models within interactive literary simulations. This system moves beyond static persona imitation or isolated scene generation by modeling literary simulation as a long-horizon process where persistent updates to character and world states occur through ongoing interactions. EvolvingWorld utilizes an open-schema approach, featuring a Character Agent for role-play and profile evolution, and an LLM-based World Model for state maintenance and scene progression. The framework is supported by a dataset derived from 57 books, comprising over 138,000 training samples and a comprehensive evaluation protocol. AI
IMPACT This framework could advance the development of more dynamic and persistent AI characters in interactive storytelling and simulation environments.
RANK_REASON The cluster describes a new research framework and benchmark published in an academic paper.
Read on Hugging Face Daily Papers →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
