New AI Research Presents AttrPrompt: An LLM as a Training Data Generator for a New Paradigm in Zero-Shot Learning
https://arxiv.org/abs/2306.15895 The performance of large language models (LLM) has been impressive in many different natural language processing (NLP) applications. In recent studies, LLMs have been proposed as generators of task-specific training data to reduce the need for task-specific data and annotations, particularly for text classification. While these efforts have demonstrated the utility of LLMs as … Read more