A Lightweight Multi Aspect Controlled Text Generation Solution For Large Language Models

TL;DR

To activate MCTG ability of LLMs, this work proposes a lightweight MCTG pipeline based on data augmentation, and analyzes bias and correlations in traditional datasets, and addresses concerns with augmented control attributes and sentences.

Abstract

Large language models (LLMs) show remarkable abilities with instruction tuning. However, they fail to achieve ideal tasks when lacking high-quality instruction tuning data on target tasks. Multi-Aspect Controllable Text Generation (MCTG) is a representative task for this dilemma, where aspect datasets are usually biased and correlated. Existing work exploits additional model structures and strategies for solutions, limiting adaptability to LLMs. To activate MCTG ability of LLMs, we propose a lightweight MCTG pipeline based on data augmentation. We analyze bias and correlations in traditional datasets, and address these concerns with augmented control attributes and sentences. Augmented datasets are feasible for instruction tuning. In our experiments, LLMs perform better in MCTG after data augmentation, with a 20% accuracy rise and less aspect correlations.

Figures (6)

• Figure 1: An overview of our lightweight MCTG solution.
• Figure 2: The prompt of Aspect-Cross Augmentation
• Figure 3: The prompt of Aspect-Grained Augmentation
• Figure 4: The prompt of Aspect-Rewrite Augmentation
• Figure 5: An instance of instruction datasets for MCTG.