Memorizing is Not Enough: Deep Knowledge Injection Through Reasoning
This work addresses the need for more effective knowledge injection in LLMs beyond memorization, though it is incremental in nature.
The paper tackles the problem of outdated or domain-specific knowledge in large language models by proposing a four-tier framework for knowledge injection, and introduces a synthetic testbed to evaluate injection depth across different knowledge types, revealing key factors and mapping injection methods to levels.
Although large language models (LLMs) excel in knowledge recall and reasoning, their static nature leads to outdated information as the real world evolves or when adapting to domain-specific knowledge, highlighting the need for effective knowledge injection. However, current research on knowledge injection remains superficial, mainly focusing on knowledge memorization and retrieval. This paper proposes a four-tier knowledge injection framework that systematically defines the levels of knowledge injection: memorization, retrieval, reasoning, and association. Based on this framework, we introduce DeepKnowledge, a synthetic experimental testbed designed for fine-grained evaluation of the depth of knowledge injection across three knowledge types (novel, incremental, and updated). We then explore various knowledge injection scenarios and evaluate the depth of knowledge injection for each scenario on the benchmark. Experimental results reveal key factors to reach each level of knowledge injection for LLMs and establish a mapping between the levels of knowledge injection and the corresponding suitable injection methods, aiming to provide a comprehensive approach for efficient knowledge injection across various levels.