BiT-MCTS: A Theme-based Bidirectional MCTS Approach to Chinese Fiction Generation

TL;DR

BiT-MCTS is a novel theme-driven framework that uses a bidirectional MCTS approach to generate long, coherent Chinese fiction with improved structure and thematic depth.

Contribution

It introduces a climax-first, bidirectional expansion strategy guided by Freytag's Pyramid, enhancing narrative coherence and diversity in large language model-generated stories.

Findings

Improves narrative coherence and plot structure.

Enables longer, more coherent stories.

Outperforms strong baselines in automatic and human evaluations.

Abstract

Generating long-form linear fiction from open-ended themes remains a major challenge for large language models, which frequently fail to guarantee global structure and narrative diversity when using premise-based or linear outlining approaches. We present BiT-MCTS, a theme-driven framework that operationalizes a "climax-first, bidirectional expansion" strategy motivated by Freytag's Pyramid. Given a theme, our method extracts a core dramatic conflict and generates an explicit climax, then employs a bidirectional Monte Carlo Tree Search (MCTS) to expand the plot backward (rising action, exposition) and forward (falling action, resolution) to produce a structured outline. A final generation stage realizes a complete narrative from the refined outline. We construct a Chinese theme corpus for evaluation and conduct extensive experiments across three contemporary LLM backbones. Results show that BiT-MCTS improves narrative coherence, plot structure, and thematic depth relative to strong baselines, while enabling substantially longer, more coherent stories according to automatic metrics and human judgments.