Saved in:
Bibliographic Details
Main Authors: Xiao, Xiongye, Ping, Heng, Zhou, Chenyu, Cao, Defu, Li, Yaxing, Zhou, Yi-Zhuo, Li, Shixuan, Kanakaris, Nikos, Bogdan, Paul
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.09099
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912521718333440
author Xiao, Xiongye
Ping, Heng
Zhou, Chenyu
Cao, Defu
Li, Yaxing
Zhou, Yi-Zhuo
Li, Shixuan
Kanakaris, Nikos
Bogdan, Paul
author_facet Xiao, Xiongye
Ping, Heng
Zhou, Chenyu
Cao, Defu
Li, Yaxing
Zhou, Yi-Zhuo
Li, Shixuan
Kanakaris, Nikos
Bogdan, Paul
contents In recent years, there has been increasing attention on the capabilities of large models, particularly in handling complex tasks that small-scale models are unable to perform. Notably, large language models (LLMs) have demonstrated ``intelligent'' abilities such as complex reasoning and abstract language comprehension, reflecting cognitive-like behaviors. However, current research on emergent abilities in large models predominantly focuses on the relationship between model performance and size, leaving a significant gap in the systematic quantitative analysis of the internal structures and mechanisms driving these emergent abilities. Drawing inspiration from neuroscience research on brain network structure and self-organization, we propose (i) a general network representation of large models, (ii) a new analytical framework, called Neuron-based Multifractal Analysis (NeuroMFA), for structural analysis, and (iii) a novel structure-based metric as a proxy for emergent abilities of large models. By linking structural features to the capabilities of large models, NeuroMFA provides a quantitative framework for analyzing emergent phenomena in large models. Our experiments show that the proposed method yields a comprehensive measure of network's evolving heterogeneity and organization, offering theoretical foundations and a new perspective for investigating emergent abilities in large models.
format Preprint
id arxiv_https___arxiv_org_abs_2402_09099
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Neuron-based Multifractal Analysis of Neuron Interaction Dynamics in Large Models
Xiao, Xiongye
Ping, Heng
Zhou, Chenyu
Cao, Defu
Li, Yaxing
Zhou, Yi-Zhuo
Li, Shixuan
Kanakaris, Nikos
Bogdan, Paul
Artificial Intelligence
In recent years, there has been increasing attention on the capabilities of large models, particularly in handling complex tasks that small-scale models are unable to perform. Notably, large language models (LLMs) have demonstrated ``intelligent'' abilities such as complex reasoning and abstract language comprehension, reflecting cognitive-like behaviors. However, current research on emergent abilities in large models predominantly focuses on the relationship between model performance and size, leaving a significant gap in the systematic quantitative analysis of the internal structures and mechanisms driving these emergent abilities. Drawing inspiration from neuroscience research on brain network structure and self-organization, we propose (i) a general network representation of large models, (ii) a new analytical framework, called Neuron-based Multifractal Analysis (NeuroMFA), for structural analysis, and (iii) a novel structure-based metric as a proxy for emergent abilities of large models. By linking structural features to the capabilities of large models, NeuroMFA provides a quantitative framework for analyzing emergent phenomena in large models. Our experiments show that the proposed method yields a comprehensive measure of network's evolving heterogeneity and organization, offering theoretical foundations and a new perspective for investigating emergent abilities in large models.
title Neuron-based Multifractal Analysis of Neuron Interaction Dynamics in Large Models
topic Artificial Intelligence
url https://arxiv.org/abs/2402.09099