This article is a series of LLM articles, focusing on the translation of "Trends in Integration of Knowledge and Large Language Models: A Survey and Taxonomy of Methods, Benchmarks, and Applications".
Trends in integrating knowledge with large language models: A survey and classification of methods, benchmarks, and applications
Summary
Large language models (LLMs) have shown excellent performance on a variety of natural language tasks, but they are susceptible to problems caused by outdated data and domain-specific limitations. To address these challenges, researchers have adopted two main strategies, namely knowledge editing and retrieval enhancement, to enhance LLM by integrating external information from different aspects. Still, there's a clear lack of a comprehensive investigation. In this paper, we present a survey to discuss trends in knowledge and large language model integration, including classification of methods, benchmarks, and applications. Furthermore, we provide an in-depth analysis of different approaches and point out potential future research directions. We hope this survey will provide the community with quick access and a comprehensive overview of this research area to inspire future research efforts.
1 Introduction
2 Knowledge editor
3 Search enhancement
4 Preface Application
5 future directions
6 Conclusion
In this paper, we survey the integration of knowledge and large language models and provide a broad view of its main directions, including knowledge editing and retrieval enhancement. Furthermore, we summarize commonly used benchmarks and cutting-edge applications and point out some promising research directions. We hope this survey will provide readers with a clear understanding of current progress and inspire additional work.