General LLMs as Instructors for Domain-Specific LLMs: A Sequential Fusion Method to Integrate Extraction and Editing

View a PDF of the paper titled General LLMs as Instructors for Domain-Specific LLMs: A Sequential Fusion Method to Integrate Extraction and Editing, by Xin Zhang and 4 other authors

View PDF
HTML (experimental)

Abstract:The substantial interest in updating Large Language Models (LLMs) without retraining from scratch is accompanied by several challenges. This is particularly true when updating LLMs with datasets that necessitate domain-expert reasoning across extensive texts, despite limited samples. We termed the scenario as the Few-Shot Domain-Expert Reasoning for Updating LLMs (FDoR-UL). Traditional methods such as Low-Rank Adaptation (LoRA) and Retrieval Augmented Generation (RAG) are inadequate for addressing this critical issue, particularly evident in our exploration of a specific medical dataset that epitomizes the distinct needs of FDoR-UL. To tackle this challenge, we introduce a Sequential Fusion method to integrate knowledge from complex contexts into LLMs. This method employs a two-stage framework: initially leveraging general LLMs to perform relation extraction for knowledge acquisition from complex texts, followed by updating domain-specific LLMs through Knowledge Editing (KE). Employing our method, domain-specific LLMs achieved a 71.7% accuracy (an average gain of 39.1%) in question-answering tasks. Furthermore, we expanded our evaluation to a novel economics-management dataset we developed, where our method achieved a 75.0% accuracy (an average gain of 45.0%). These findings underscore the effectiveness and flexibility of our approach in FDoR-UL across various domains.