Customizing Large Language Model Generation Style using Parameter-Efficient Finetuning

TL;DR

This paper investigates using parameter-efficient finetuning with Low-Rank Adaptation to customize large language model styles to match individual authors, enhancing user-specific writing assistance.

Contribution

It demonstrates that PEFT can effectively adapt LLMs to individual writing styles with lexical and syntactic alignment, offering a practical approach for user customization.

Findings

Lexical and syntactic style matching achieved

Content memorization remains a challenge

PEFT enables efficient style customization

Abstract

One-size-fits-all large language models (LLMs) are increasingly being used to help people with their writing. However, the style these models are trained to write in may not suit all users or use cases. LLMs would be more useful as writing assistants if their idiolect could be customized to match each user. In this paper, we explore whether parameter-efficient finetuning (PEFT) with Low-Rank Adaptation can effectively guide the style of LLM generations. We use this method to customize LLaMA-2 to ten different authors and show that the generated text has lexical, syntactic, and surface alignment with the target author but struggles with content memorization. Our findings highlight the potential of PEFT to support efficient, user-level customization of LLMs.