Pruning Large Language Models by Identifying and Preserving Functional Networks

TL;DR

This paper introduces a novel structured pruning method for large language models that preserves functional neural networks, maintaining model performance while reducing size and computational requirements.

Contribution

It proposes a new approach to pruning LLMs by identifying and preserving functional networks, inspired by brain neural network analysis, improving pruning effectiveness.

Findings

Successfully identifies functional networks in LLMs

Preserves key neurons within networks during pruning

Achieves efficient model compression with maintained performance

Abstract

Structured pruning is one of the representative techniques for compressing large language models (LLMs) to reduce GPU memory consumption and accelerate inference speed. It offers significant practical value in improving the efficiency of LLMs in real-world applications. Current structured pruning methods typically rely on assessment of the importance of the structure units and pruning the units with less importance. Most of them overlooks the interaction and collaboration among artificial neurons that are crucial for the functionalities of LLMs, leading to a disruption in the macro functional architecture of LLMs and consequently a pruning performance degradation. Inspired by the inherent similarities between artificial neural networks and functional neural networks in the human brain, we alleviate this challenge and propose to prune LLMs by identifying and preserving functional networks within LLMs in this study. To achieve this, we treat an LLM as a digital brain and decompose the LLM into functional networks, analogous to identifying functional brain networks in neuroimaging data. Afterwards, an LLM is pruned by preserving the key neurons within these functional networks. Experimental results demonstrate that the proposed method can successfully identify and locate functional networks and key neurons in LLMs, enabling efficient model pruning. Our code is available at https://github.com/WhatAboutMyStar/LLM_ACTIVATION.