Scaling Creative Writing Beyond Story-Centric Data with Attribute-Guided Genre Expansion

• 类型：arxiv
• 标识：2608.13947
• 链接：https://arxiv.org/abs/2608.13947
• 主分类：llm-infra
• 形态：method
• 被引：0
• 被引来源：Semantic Scholar
• S2被引：0
• 影响力被引：0
• TLDR：Experiments demonstrate that models fine-tuned on the data consistently surpass not only base models and writing-specialized baselines, but also models trained on existing writing corpora, indicating that controlled genre expansion is a key driver of robust creative writing capability.
• 待LLM分类：否
• 成熟度：research
• 场景：creative writing、fine-tuning、data augmentation
• 标题中文：超越故事中心数据的可扩展创意写作：基于属性引导的题材扩展
• TLDR中文：实验表明，在该数据上微调的模型不仅持续优于基座模型和面向写作的专门基线，还优于基于现有写作语料训练的模型，表明受控的题材扩展是稳健创意写作能力的关键驱动力
• 来源文件：
• /inbox/tom/_candidates/2026-08-20-agent-rag-longcontext-candidates.json
• [S2 enrich]