The Rise of Parameter Specialization for Knowledge Storage in Large Language Models

TL;DR

This paper investigates how large language models store knowledge in their MLP parameters, revealing that increased specialization of parameters correlates with improved knowledge utilization and model performance.

Contribution

It provides the first detailed analysis of knowledge storage in MLP parameters of large language models and demonstrates the importance of parameter specialization for efficient knowledge use.

Findings

Parameters in MLPs become more specialized in advanced models.

Specialized knowledge distribution enhances model efficiency.

Causal experiments confirm the role of specialization in knowledge utilization.

Abstract

Over time, a growing wave of large language models from various series has been introduced to the community. Researchers are striving to maximize the performance of language models with constrained parameter sizes. However, from a microscopic perspective, there has been limited research on how to better store knowledge in model parameters, particularly within MLPs, to enable more effective utilization of this knowledge by the model. In this work, we analyze twenty publicly available open-source large language models to investigate the relationship between their strong performance and the way knowledge is stored in their corresponding MLP parameters. Our findings reveal that as language models become more advanced and demonstrate stronger knowledge capabilities, their parameters exhibit increased specialization. Specifically, parameters in the MLPs tend to be more focused on encoding similar types of knowledge. We experimentally validate that this specialized distribution of knowledge contributes to improving the efficiency of knowledge utilization in these models. Furthermore, by conducting causal training experiments, we confirm that this specialized knowledge distribution plays a critical role in improving the model's efficiency in leveraging stored knowledge.