arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2502.08356cs.CL
收藏

通过多样化增强向大型语言模型中系统性注入领域特定RAG知识

Systematic Knowledge Injection into Large Language Models via Diverse Augmentation for Domain-Specific RAG

IIT(ISM) Dhanbad(丹巴印度理工学院(印度矿业学院));IBM Research(IBM研究院)

Kushagra Bhushan, Yatin Nandwani, Dinesh Khandelwal, Sonam Gupta, Gaurav Pandey, Dinesh Raghu, Sachindra Joshi

AI总结:

提出通过上下文增强和知识释义的系统性数据增强框架,改进领域特定RAG中LLM的微调,提升知识注入效果并缓解灾难性遗忘。

摘要:

检索增强生成(RAG)已成为将领域知识融入大型语言模型(LLM)的一种重要方法。尽管RAG通过将检索到的领域知识纳入上下文来增强回答的相关性,但检索错误仍可能导致幻觉和错误答案。为了从检索器故障中恢复,通过微调模型以在检索错误的情况下也能生成正确响应,从而注入领域知识。然而,我们观察到,如果没有系统性的知识增强,微调后的LLM可能记住新信息,但仍无法提取相关的领域知识,导致性能不佳。在这项工作中,我们提出了一种新颖的框架,通过两种方式增强训练数据来显著改进微调过程——上下文增强和知识释义。在上下文增强中,我们通过改变检索信息的相关性为给定的问答对创建多个训练样本,教导模型何时忽略以及何时依赖检索到的内容。在知识释义中,我们使用同一问题的多个答案进行微调,使LLM能够更好地内化专业知识。为了减轻微调导致的灾难性遗忘,我们在问题中添加领域特定标识符,并利用包含通用问答对的回放缓冲区。实验结果表明,我们的方法优于现有技术,在保持LLM泛化能力的同时,实现了高达10%的令牌级召回率相对提升。

英文摘要:

Retrieval-Augmented Generation (RAG) has emerged as a prominent method for incorporating domain knowledge into Large Language Models (LLMs). While RAG enhances response relevance by incorporating retrieved domain knowledge in the context, retrieval errors can still lead to hallucinations and incorrect answers. To recover from retriever failures, domain knowledge is injected by fine-tuning the model to generate the correct response, even in the case of retrieval errors. However, we observe that without systematic knowledge augmentation, fine-tuned LLMs may memorize new information but still fail to extract relevant domain knowledge, leading to poor performance. In this work, we present a novel framework that significantly enhances the fine-tuning process by augmenting the training data in two ways -- context augmentation and knowledge paraphrasing. In context augmentation, we create multiple training samples for a given QA pair by varying the relevance of the retrieved information, teaching the model when to ignore and when to rely on retrieved content. In knowledge paraphrasing, we fine-tune with multiple answers to the same question, enabling LLMs to better internalize specialized knowledge. To mitigate catastrophic forgetting due to fine-tuning, we add a domain-specific identifier to a question and also utilize a replay buffer containing general QA pairs. Experimental results demonstrate the efficacy of our method over existing techniques, achieving up to 10\% relative gain in token-level recall while preserving the LLM's generalization capabilities.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑