Saved in:
Bibliographic Details
Main Authors: Liu, Jing, Wang, Haozheng, Li, Yueheng
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.12822
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910962611650560
author Liu, Jing
Wang, Haozheng
Li, Yueheng
author_facet Liu, Jing
Wang, Haozheng
Li, Yueheng
contents Large language models struggle with representing and generating rare tokens despite their importance in specialized domains. In this study, we identify neuron structures with exceptionally strong influence on language model's prediction of rare tokens, termed as rare token neurons, and investigate the mechanism for their emergence and behavior. These neurons exhibit a characteristic three-phase organization (plateau, power-law, and rapid decay) that emerges dynamically during training, evolving from a homogeneous initial state to a functionally differentiated architecture. In the activation space, rare token neurons form a coordinated subnetwork that selectively co-activates while avoiding co-activation with other neurons. This functional specialization potentially correlates with the development of heavy-tailed weight distributions, suggesting a statistical mechanical basis for emergent specialization.
format Preprint
id arxiv_https___arxiv_org_abs_2505_12822
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Emergent Specialization: Rare Token Neurons in Language Models
Liu, Jing
Wang, Haozheng
Li, Yueheng
Artificial Intelligence
Large language models struggle with representing and generating rare tokens despite their importance in specialized domains. In this study, we identify neuron structures with exceptionally strong influence on language model's prediction of rare tokens, termed as rare token neurons, and investigate the mechanism for their emergence and behavior. These neurons exhibit a characteristic three-phase organization (plateau, power-law, and rapid decay) that emerges dynamically during training, evolving from a homogeneous initial state to a functionally differentiated architecture. In the activation space, rare token neurons form a coordinated subnetwork that selectively co-activates while avoiding co-activation with other neurons. This functional specialization potentially correlates with the development of heavy-tailed weight distributions, suggesting a statistical mechanical basis for emergent specialization.
title Emergent Specialization: Rare Token Neurons in Language Models
topic Artificial Intelligence
url https://arxiv.org/abs/2505.12822