Get Started
Home
Topics
Search
Library
Research questionHow can interactive literary simulations preserve coherent character and world evolution over long horizons?As interactions accumulate, character behavior, scene progression, and persistent world state can drift out of alignment. This makes it difficult to sustain a consistent evolving literary world rather than isolated plausible scenes.
AI
AI Agents
Evaluation & Benchmarks
Multi-agent Systems
Natural Language Processing
Latest papersRecent research connected to this question, newest first.ArcANE: Do Role-Playing Language Agents Stay in Character at the Right Time?ArcANE evaluates six models on scenes from a novel, new situations within its world, and situations outside it, under six forms of narrative context. Its reported evidence indicates that providing the character arc through the queried chapter improves performance over the strongest non-arc context by 2.2–8.4 points; the evidence is limited to the evaluated models and scenarios.research paper · Sep 2, 2026EvolvingWorld: An Open-Schema Framework for Co-Evolving Role-Play Agents and World Model in Interactive Literary WorldThe evidence concerns EvolvingWorld’s open-schema framework and benchmark, which couple a Character Agent with an LLM-based World Model. It covers seven trainable tasks, data from 57 books, 138,596 training samples, 222 test snapshots, and trajectory-level LLM-as-Judge evaluation across 10 dimensions and 20 metrics; the source does not establish how well the approach transfers beyond these literary-world data.research paper · Jul 19, 2026
Related questions
How can text-based world models be evaluated for behavioral fidelity in long-horizon agent planning?How can language agents adapt textual world models to evolving behavior in interactive environments without external rewards?How can multimodal language-model agents coordinate hidden prerequisites during long-horizon open-world exploration?How can agent runtimes avoid context poisoning and latency from growing histories during long-horizon skill execution?