Synchronization of mean-field models on the circle

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Polyanskiy, Yury, Rigollet, Philippe, Yao, Andrew
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918107838152704
author Polyanskiy, Yury
Rigollet, Philippe
Yao, Andrew
author_facet Polyanskiy, Yury
Rigollet, Philippe
Yao, Andrew
contents This paper considers a mean-field model of $n$ interacting particles whose state space is the unit circle, a generalization of the classical Kuramoto model. Global synchronization is said to occur if after starting from almost any initial state, all particles coalesce to a common point on the circle. We propose a general synchronization criterion in terms of $L_1$-norm of the third derivative of the particle interaction function. As an application we resolve a conjecture for the so-called self-attention dynamics (stylized model of transformers), by showing synchronization for all $β\ge -0.16$, which significantly extends the previous bound of $0\le β\le 1$ from Criscitiello, Rebjock, McRae, and Boumal (2024). We also show that global synchronization does not occur when $β< -2/3$.
format Preprint
id arxiv_https___arxiv_org_abs_2507_22857
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Synchronization of mean-field models on the circle
Polyanskiy, Yury
Rigollet, Philippe
Yao, Andrew
Dynamical Systems
Machine Learning
Analysis of PDEs
Optimization and Control
This paper considers a mean-field model of $n$ interacting particles whose state space is the unit circle, a generalization of the classical Kuramoto model. Global synchronization is said to occur if after starting from almost any initial state, all particles coalesce to a common point on the circle. We propose a general synchronization criterion in terms of $L_1$-norm of the third derivative of the particle interaction function. As an application we resolve a conjecture for the so-called self-attention dynamics (stylized model of transformers), by showing synchronization for all $β\ge -0.16$, which significantly extends the previous bound of $0\le β\le 1$ from Criscitiello, Rebjock, McRae, and Boumal (2024). We also show that global synchronization does not occur when $β< -2/3$.
title Synchronization of mean-field models on the circle
topic Dynamical Systems
Machine Learning
Analysis of PDEs
Optimization and Control
url https://arxiv.org/abs/2507.22857