Guardado en:
Detalles Bibliográficos
Autores principales: Sun, Youran, Haghighat, Babak
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2501.16241
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916585564798976
author Sun, Youran
Haghighat, Babak
author_facet Sun, Youran
Haghighat, Babak
contents Large language models (LLMs) exhibit unprecedentedly rich scaling behaviors. In physics, scaling behavior is closely related to phase transitions, critical phenomena, and field theory. To investigate the phase transition phenomena in LLMs, we reformulated the Transformer architecture as an $O(N)$ model. Our study reveals two distinct phase transitions corresponding to the temperature used in text generation and the model's parameter size, respectively. The first phase transition enables us to estimate the internal dimension of the model, while the second phase transition is of \textit{higher-depth} and signals the emergence of new capabilities. As an application, the energy of the $O(N)$ model can be used to evaluate whether an LLM's parameters are sufficient to learn the training data.
format Preprint
id arxiv_https___arxiv_org_abs_2501_16241
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Phase Transitions in Large Language Models and the $O(N)$ Model
Sun, Youran
Haghighat, Babak
Machine Learning
Computation and Language
High Energy Physics - Theory
Data Analysis, Statistics and Probability
Large language models (LLMs) exhibit unprecedentedly rich scaling behaviors. In physics, scaling behavior is closely related to phase transitions, critical phenomena, and field theory. To investigate the phase transition phenomena in LLMs, we reformulated the Transformer architecture as an $O(N)$ model. Our study reveals two distinct phase transitions corresponding to the temperature used in text generation and the model's parameter size, respectively. The first phase transition enables us to estimate the internal dimension of the model, while the second phase transition is of \textit{higher-depth} and signals the emergence of new capabilities. As an application, the energy of the $O(N)$ model can be used to evaluate whether an LLM's parameters are sufficient to learn the training data.
title Phase Transitions in Large Language Models and the $O(N)$ Model
topic Machine Learning
Computation and Language
High Energy Physics - Theory
Data Analysis, Statistics and Probability
url https://arxiv.org/abs/2501.16241