Mesoscopic theory of flocking with alignment and anti-alignment copying

Chunming Zheng · 2026

Read Mesoscopic theory of flocking with alignment and anti-alignment copying by Chunming Zheng (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-20251.txt.

Introduction

Zheng studies flocking where members copy a neighbor's heading, lining up or against it. Dice-roll dynamics redraw copy types each step; fixed dynamics lock members in lining-up or lining-against groups. From a microscopic master equation on the circle, Fourier and large-number expansions yield Fokker-Planck equations and random equations for polarization. Rival lining up and lining against wipe out long-range polar order in huge systems, while finite flocks show noise-driven structure set by the mix, as Gillespie runs confirm.

Catalog record

CREATORChunming Zheng
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE35626
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.20251

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Mesoscopic theory of flocking with alignment and anti-alignment copying Chunming Zheng Abstract We study a stochastic model of collective motion in which individuals update their orientation through pairwise aligning or anti-aligning copying interactions. We analyze both annealed dynamics, where interaction types are chosen probabilistically at each update, and quenched dynamics, where individuals are permanently assigned to aligning or anti-aligning subpopulations. Starting from the microscopic master equation on the circle, we derive an exact mesoscopic description via a Fourier-mode expansion and a systematic large N N expansion, obtaining closed Fokker–Planck equations and effective stochastic differential equations for the polarization. We show that competing alignment and anti-alignment suppress long-range polar order in the thermodynamic limit in both cases, while finite systems display nontrivial fluctuation-induced structure controlled by the interaction composition. Our results, validated by Gillespie simulations, establish an analytically tractable framework for collective dynamics characterized by competing copying rules and intrinsic noise. I Introduction Collective motion is one of the most striking emergent behaviors exhibited by biological systems, exemplified by the coordinated maneuvers of bird flocks and fish schools. The spontaneous formation of coherent group motion, often termed flocking or schooling, arises from simple local interactions among individuals yet displays complex global order 28 ; 26 ; 11 ; 3 ; 6 ; 2 ; 9 ; 25 ; 14 . Understanding how microscopic interaction rules give rise to mesoscopic dynamics and ultimately macroscopic organization remains a central challenge in the physics of active and living matter. While alignment-based models are often studied at the deterministic or hydrodynamic level 28 ; 26 ; 11 ; 3 , a complementary line of research formulates these interactions at the stochastic, individual level using Master equations, thereby explicitly incorporating demographic noise and enabling systematic coarse-graining to mesoscopic dynamics 4 ; 12 . Within this framework, intrinsic fluctuations arising from the discrete nature of individuals can qualitatively alter system behavior, driving states and transitions that are absent in deterministic mean-field descriptions. This perspective has motivated the derivation of mesoscopic descriptions, e.g. Fokker–Planck or Langevin equations, directly from microscopic interaction rules. Notably, the underlying interaction structure in many such models closely parallels the voter model, a paradigmatic stochastic process in which agents update their states via imitation or copying 23 ; 8 . Building on this connection, Jhawar et al. 20 showed experimentally that finite-size fluctuations alone can induce spontaneous polarization even in the absence of deterministic bias, highlighting the constructive role of noise in the emergent collective order. Beyond alignment mechanisms, collective systems can also exhibit interactions that favor opposite orientations, resulting in effective anti-alignment of velocity directions. Such anti-alignment or repulsive interactions have been explored in active matter models in both single- and multi-population settings 18 ; 29 ; 10 ; 13 ; 21 ; 22 . In contrast, within stochastic, reaction-like frameworks, where interactions occur through probabilistic copying of orientations, the interplay between alignment and anti-alignment interactions remains largely unexplored. In this paper, we introduce and analyze a minimal stochastic model of collective orientation dynamics driven by competing alignment and anti-alignment interactions, together with spontaneous random turning. Starting from the microscopic Master equation, we derive exact mesoscopic descriptions for the polarization order parameter in the form of Fokker–Planck and Langevin equations. Our formulation can be viewed as a vector generalization of the voter model: whereas the standard voter model describes the imitation of discrete opinions 23 ; 8 , our model extends these rules to continuous angular variables with circular symmetry, incorporating both alignment and “anti-voter” (anti-alignment) interactions. We consider two realizations of the interaction structure. In the annealed version, interactions are transient: an individual adopts the direction of a randomly selected partner with probability p (alignment) or the opposite direction with probability 1-p (anti-alignment). In the quenched version, the population is heterogeneous: a fixed fraction p of individuals are aligners who always follow a randomly selected partner, while the remaining 1-p are anti-aligners who adopt the opposite direction. This quenched heterogeneity is closely related to the conformist–contrarian framework introduced in coupled oscillators 19 and to the ’anti-voter’ and non-conformist behaviors studied in social dynamics 16 ; 8 , as well as the competing excitatory–inhibitory interactions fundamental to neural network models 5 . The mixed copying model considered here generalizes the purely aligning case studied in Ref. 20 and provides a tractable framework to investigate how cooperative and antagonistic interactions jointly shape the effective drift and noise governing collective order. II Annealed alignment and anti-alignment copying In the annealed realization, we consider a system of N agents, where each agent i\in\{1,\dots,N\} is characterized by a continuous orientation \theta_{i}\in[-\pi,\pi) . The state of the system evolves through two stochastic processes. First, spontaneous random turning occurs at a rate a , where an agent i resets its orientation to a new value \theta^{*} drawn from a uniform distribution \mathcal{U}(-\pi,\pi) . This process ensures rotational symmetry in the absence of interactions. Second, pairwise interactions occur at a total rate c . In an interaction event, a ’focal’ agent i and a ’partner’ agent j are selected; agent i then adopts the orientation of j with probability p (alignment) or the opposite direction (\theta_{j}+\pi)\pmod{2\pi} with probability 1-p (anti-alignment). These rules are summarized by the following reaction scheme: \begin{cases}\theta_{i}\xrightarrow{a}\theta^{*}\in\mathcal{U}(-\pi,\pi)&\text{(Random turning)}\\ \theta_{i}+\theta_{j}\xrightarrow{cp}2\theta_{j}&\text{(Alignment copying)}\\ \theta_{i}+\theta_{j}\xrightarrow{c(1-p)}(\theta_{j}+\pi)+\theta_{j}&\text{(Anti-alignment copying)}.\end{cases} (1) Numerical simulations of these microscopic processes are implemented using the Gillespie stochastic simulation algorithm 17 , which provides an exact realization of the underlying Master equation. To obtain a mesoscopic description, we introduce the empirical orientation density \varphi(x)=\frac{1}{N}\sum_{i=1}^{N}\delta(x-\theta_{i}), (2) which represents the fraction of individuals with orientation x . To analyze the collective dynamics, we decompose the empirical density into a Fourier series, \varphi(x)=\frac{1}{2\pi}\sum_{k}\varphi_{k}e^{-ikx} . The complex amplitudes \varphi_{k}=\int_{-\pi}^{\pi}\varphi(x)e^{ikx}dx serve as the fundamental order parameters, where \varphi_{0}=1 ensures normalization and \varphi_{1} corresponds to the global polar order vector \mathbf{m} . The stochastic dynamics of the microscopic reactions induces a Markovian evolution for the probability functional P(\varphi,t) of observing a density configuration \varphi(x) at time t . Following standard reaction-process methods 27 , the corresponding functional master equation can be written as \frac{d}{dt}P(\varphi,t)=N\int_{-\pi}^{\pi}dx\int_{-\pi}^{\pi}dy\,\mathcal{Q}(\varphi;x,y)P(\varphi,t), (3) where the transition operator \mathcal{Q} encapsulates the local field dynamics: \mathcal{Q}=(\Delta_{x}^{+}\Delta_{y}^{-}-1)\varphi(x)\left[\frac{a}{2\pi}+c\gamma(\varphi,y)\right]. (4) Here, \gamma(\varphi,y)=p\varphi(y)+(1-p)\varphi(y-\pi) accounts for the mixed alignment ( p ) and anti-alignment ( 1-p ) copying interactions. The step operators \Delta_{x}^{\pm} denote the addition or removal of a particle at orientation x . In the large N limit, we expand it to the second order in 1/N 24 : \Delta_{x}^{\pm}\approx 1\pm\frac{1}{N}\sum_{k}e^{-ikx}\partial_{k}+\frac{1}{2N^{2}}\sum_{k,\ell}e^{-i(k+\ell)x}\partial_{k}\partial_{\ell}, (5) where \partial_{k}\equiv\delta/\delta\varphi_{k} . Substituting this expansion and the Fourier decomposition \varphi(x)=\frac{1}{2\pi}\sum_{k}\varphi_{k}e^{-ikx} into Eq. ( 3 ), the spatial integrals over x and y are evaluated. Due to the orthogonality of the Fourier modes, \int_{-\pi}^{\pi}e^{i(k-m)x}dx=2\pi\delta_{k,m} , the continuous field evolution collapses into a multivariate Fokker-Planck equation for the joint distribution of modes P(\{\varphi_{k}\},t) : \partial_{t}P=-\sum_{k}\partial_{k}(A_{k}P)+\sum_{k,\ell}\partial_{k}\partial_{\ell}(B_{k\ell}P). (6) The drift and diffusion coefficients are derived as: \displaystyle A_{k} \displaystyle=-[a+c(1-\Gamma_{k})]\varphi_{k}, (7) \displaystyle B_{k\ell} \displaystyle=\frac{1}{2N}\left[a\varphi_{k+\ell}+c\left((1+(-1)^{k+\ell})\varphi_{k+\ell}-2\Gamma_{\ell}\varphi_{k}\varphi_{\ell}\right)\right], where we define the coupling weight \Gamma_{k}=p+(1-p)(-1)^{k} . For the polar mode ( k=1 ), the coupling weight simplifies to \Gamma_{1}=2p-1 , which directly governs the onset of collective order as a function of the alignment probability p . II.1 Polar dynamics and noise-induced ordering Figure 1: Annealed dynamics results. (a) Probability density functions (PDFs) of the

Read the complete text