RET-LLM: Towards a General Read-Write Memory for Large Language Models
This addresses the limitation of explicit knowledge storage and retrieval in LLMs for NLP tasks, though it appears incremental as it builds on existing LLM architectures.
The authors tackled the problem of LLMs lacking a dedicated memory unit by proposing RET-LLM, a framework that adds a general write-read memory unit to store and retrieve knowledge as triplets, showing superiority in question answering tasks and robust performance in temporal-based tasks.
Large language models (LLMs) have significantly advanced the field of natural language processing (NLP) through their extensive parameters and comprehensive data utilization. However, existing LLMs lack a dedicated memory unit, limiting their ability to explicitly store and retrieve knowledge for various tasks. In this paper, we propose RET-LLM a novel framework that equips LLMs with a general write-read memory unit, allowing them to extract, store, and recall knowledge from the text as needed for task performance. Inspired by Davidsonian semantics theory, we extract and save knowledge in the form of triplets. The memory unit is designed to be scalable, aggregatable, updatable, and interpretable. Through qualitative evaluations, we demonstrate the superiority of our proposed framework over baseline approaches in question answering tasks. Moreover, our framework exhibits robust performance in handling temporal-based question answering tasks, showcasing its ability to effectively manage time-dependent information.