Meta-Tuning LLMs to Leverage Lexical Knowledge for Generalizable Language Style Understanding
This addresses the challenge of generalizable language style understanding for writers and NLP applications, though it is incremental as it builds on existing meta-training and lexicon-based methods.
The paper tackles the problem of large language models struggling to capture language styles without fine-tuning by meta-training them with style lexicons, resulting in improved zero-shot transfer across 13 established and 63 novel style classification tasks.
Language style is often used by writers to convey their intentions, identities, and mastery of language. In this paper, we show that current large language models struggle to capture some language styles without fine-tuning. To address this challenge, we investigate whether LLMs can be meta-trained based on representative lexicons to recognize new styles they have not been fine-tuned on. Experiments on 13 established style classification tasks, as well as 63 novel tasks generated using LLMs, demonstrate that meta-training with style lexicons consistently improves zero-shot transfer across styles. We release the code and data at http://github.com/octaviaguo/Style-LLM .