1 paper
Hwan Chang, Yongil Kim, Heuiyeen Yeen +3
High-quality creative writing data for large language models (LLMs) remains dominated by story-centric data, limiting models' ability to follow the structural and functional conven…