Structured Style-Rewrite with Chain-of-Thought Planning for Low-Resource Character Dialogue

TL;DR

This paper introduces a structured style-rewrite framework with chain-of-thought supervision for low-resource Chinese character dialogue generation, improving style control and semantic fidelity in small language models.

Contribution

It proposes a novel decomposed style representation combined with chain-of-thought planning and preference optimization, enhancing style accuracy and semantic consistency.

Findings

Achieved a Valid Style Score of 0.632 on eight characters.

Maintained semantic fidelity of 0.878.

Outperformed larger baseline models on consumer hardware.

Abstract

Applying Small Language Models (SLMs) to Chinese character-driven generation remains challenging due to data scarcity and the difficulty of disentangling character style. Standard Supervised Fine-Tuning (SFT) often captures surface-level semantics but produces frequent Out-Of-Character (OOC) outputs. We frame this as a controlled sentence-level style rewriting task, which isolates stylistic quality from dialogue context management. We propose a Structured Style-Rewrite Framework that decomposes character style into interpretable format signature, syntactic, and pragmatic dimensions, combined with Chain-of-Thought (CoT) supervision for explicit style planning. A CoT-Shared Direct Preference Optimization (DPO) stage further aligns style planning with surface realization by ensuring preference learning targets output-level style execution rather than reasoning trace differences. Experiments across eight characters from four diverse source domains demonstrate that our method enables a Qwen3-1.7B model to achieve a Valid Style Score of $0.632$ while maintaining strong semantic fidelity (0.878), placing on the Pareto frontier among the evaluated systems and outperforming significantly larger baselines (e.g., GLM-4.7) on consumer hardware.