Living the Novel: A System for Generating Self-Training Timeline-Aware Conversational Agents from Novels

TL;DR

The Living Novel system transforms literary works into immersive, multi-character conversational agents by addressing persona drift and narrative incoherence through a novel two-stage training pipeline, validated on Jules Verne's novel.

Contribution

Introduces a two-stage training pipeline with Deep Persona Alignment and story-time-aware knowledge graphs for narrative coherence in literary-based conversational agents.

Findings

Outperforms GPT-4o on persona-specific metrics.

Achieves near-perfect coherence and robustness.

Provides practical guidelines for AI-driven narrative systems.

Abstract

We present the Living Novel, an end-to-end system that transforms any literary work into an immersive, multi-character conversational experience. This system is designed to solve two fundamental challenges for LLM-driven characters. Firstly, generic LLMs suffer from persona drift, often failing to stay in character. Secondly, agents often exhibit abilities that extend beyond the constraints of the story's world and logic, leading to both narrative incoherence (spoiler leakage) and robustness failures (frame-breaking). To address these challenges, we introduce a novel two-stage training pipeline. Our Deep Persona Alignment (DPA) stage uses data-free reinforcement finetuning to instill deep character fidelity. Our Coherence and Robustness Enhancing (CRE) stage then employs a story-time-aware knowledge graph and a second retrieval-grounded training pass to architecturally enforce these narrative constraints. We validate our system through a multi-phase evaluation using Jules Verne's Twenty Thousand Leagues Under the Sea. A lab study with a detailed ablation of system components is followed by a 5-day in-the-wild diary study. Our DPA pipeline helps our specialized model outperform GPT-4o on persona-specific metrics, and our CRE stage achieves near-perfect performance in coherence and robustness measures. Our study surfaces practical design guidelines for AI-driven narrative systems: we find that character-first self-training is foundational for believability, while explicit story-time constraints are crucial for sustaining coherent, interruption-resilient mobile-web experiences.