Unveiling the impact of cross-order hyperdegree correlations in contagion processes on hypergraphs

Andrés Guzmán, Federico Malizia, István Z. Kiss · 2026

Read Unveiling the impact of cross-order hyperdegree correlations in contagion processes on hypergraphs by Andrés Guzmán, Federico Malizia, István Z. Kiss (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-2601-15199.txt.

Introduction

Andrés Guzmán, Federico Malizia, and István Z. Kiss study how contagion spreads when contacts come in groups. Their 2026 arXiv paper deals with hypergraphs, networks where one link can join many people at once. Group contacts can change spread dynamics a lot.

Standard analysis treats pairwise and group links as if they were independent. The authors drop that assumption. They study cross-order hyperdegree correlations, links between a node's pairwise degree and its group degree. The work shows how those ties reshape the way a contagion moves through a whole human population.

Catalog record

CREATORAndrés Guzmán, Federico Malizia, István Z. Kiss
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE97 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-15199

Browse the full library · About the Tower of Babel

Full text

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

Unveiling the impact of cross-order hyperdegree correlations in contagion processes on hypergraphs Andrés Guzmán, Federico Malizia, István Z. Kiss Abstract Contagion processes in social systems often involve interactions that go beyond pairwise contacts. Higher-order networks, represented as hypergraphs, have been widely used to model multi-body interactions, and their presence can drastically alter contagion dynamics compared to traditional network models. However, existing analytical approaches typically assume independence between pairwise and higher-order degrees, and thus study their roles in isolation. In this paper, we develop an effective hyperdegree model (EHDM) to describe Susceptible-Infected-Susceptible (SIS) dynamics on hypergraphs that explicitly captures correlations between the distribution of groups with different sizes. Our effective hyperdegree model shows excellent agreement with stochastic simulations across different types of higher-order networks, including those with heterogeneous degree distributions. We explore the critical role of cross-order degree correlations, specifically, whether nodes that are hubs in pairwise interactions also serve as hubs in higher-order interactions. We show that positive correlation decreases the epidemic threshold and anti-correlation temporally desynchronizes infection pathways (pairwise and group interactions). Finally, we demonstrate that, depending on the level of correlation, the optimal control strategy shifts—from one that is purely pairwise- or higher-order-focused to one in which a mixed strategy becomes optimal. I Introduction Complex networks have become a fundamental framework for modeling spreading processes and understanding a wide range of phenomena, including epidemics 38 ; 47 ; 26 ; 40 ; 2 , information diffusion 13 ; 36 ; 44 , and the adoption of behaviours and social contagion 21 ; 9 ; 48 . While this line of research has produced significant advances, recent years have seen growing interest in spreading phenomena that extend beyond pairwise interactions to encompass group interactions. This framework is especially important in the context of social dynamics, as reducing group interactions to a collection of pairwise connections can miss essential features of collective behavior. Empirical evidence shows that humans exhibit markedly different interaction patterns when engaging in groups compared to dyadic encounters 18 ; 30 . Accounting for multi-body interactions has been shown to profoundly affect dynamical processes such as contagion 22 ; 14 ; 16 , synchronization 5 ; 45 ; 35 , diffusion 43 ; 8 ; 37 , and social polarization 42 . In spreading dynamics, the incorporation of higher-order interactions—i.e., interactions among groups of more than two individuals—reveals dynamical behaviors not captured by traditional pairwise network models. For reversible spreading processes such as the Susceptible–Infected–Susceptible (SIS) model, introducing three-body interactions, with links and triples representing different transmission pathways, leads to a variety of non-trivial phenomena. These include discontinuous phase transitions 22 , hysteresis loops 19 , explosive transitions 34 ; 11 , and bistability, where healthy and endemic states may coexist 22 ; 17 . More recently, considerable effort has been devoted to deepening our understanding of how structural characteristics influence what behaviors systems can exhibit. In higher-order network representations, nodes can belong to hyperedges of different cardinality, leading to not one degree but a vector of hyperdegrees. The way in which hyperdegrees are distributed is fundamental in shaping dynamical processes unfolding on higher-order networks. For example, Landry et al. showed that the characteristic explosive transition previously reported can be suppressed by heterogeneity in pairwise interactions 28 . Conversely, heterogeneity in three-body interactions has been shown to play a key role in inducing explosive transitions 32 . The richness of these systems arises not only from the incremental inclusion of higher-order interactions, but also from the interplay between interactions of different orders. Beyond hyperdegree distributions, the arrangement of hyperedges has been shown to fundamentally shape epidemic dynamics 6 ; 25 ; 24 ; 33 ; 32 . For instance, nested interactions, where lower-order interactions are embedded within higher-order ones 27 , play a dual role in shaping epidemic onset: they can anticipate the emergence of epidemics while simultaneously suppressing abrupt transitions 31 . More broadly, nestedness influences not only macroscopic properties such as the final epidemic size, but also the dynamical pathways leading to explosive transitions 32 . Indeed, the mere presence of group interactions does not guarantee abrupt transitions in SIS processes 23 ; 33 , as the structural arrangement of hyperedges, including correlations between different group sizes, profoundly shapes epidemic dynamics. However, correlations between different orders of interaction can be examined from multiple perspectives and are not limited to the presence of nested hyperedges. In many studies, the hyperdegree distributions—pairwise, three-body interactions, and higher-orders—are chosen independently from each order. This assumption is of course convenient from both algorithmic and analytical standpoints. Especially in the context of heterogeneous hyperdegree distributions, the correlations that arise when hyperdegrees are drawn from a joint distribution rather than treated as fully independent remain largely unexplored. Such correlations fundamentally alter which nodes act as hubs at different interaction orders. Landry et al 28 . developed a hypergraph generative model alongside a degree-based spreading model and showed that positive correlations between the number of pairwise and group interactions a node participates in can promote the emergence of bistability and hysteresis 28 . Nonetheless, open questions remain regarding other types of correlations—for example, cases in which pairwise and higher-order hyperdegrees are anti-correlated, which may require more flexible modeling frameworks. Moreover, it is still unclear how different patterns of degree correlation shape the onset of the epidemic, influence the hierarchy of spread and how it impacts the optimality of control strategies based on hyperdegree of nodes. In this paper, we overcome the limitations imposed by assuming independent hyperdegree distributions by introducing a configurational model with a joint distribution that naturally captures correlations between a node’s participation in groups of different sizes. Furthermore, we develop an effective hyperdegree model (EHDM) for the Susceptible–Infected–Susceptible (SIS) process that accurately captures the dynamics on heterogeneous hypergraphs while accounting for correlations between interaction orders in the hyperdegree distribution. Through a systematic analysis of the epidemic threshold, quasi-equilibrium behavior, and time evolution, we provide a comprehensive characterization of how cross-degree correlations shape the system’s dynamical properties. Finally, we investigate how contagion propagates hierarchically through nodes belonging to different hyperdegree classes in a higher-order network with both pairwise and three-body interactions. Building on these insights, we show that optimal control strategies for targeting nodes based on their hyperdegree are dictated by the underlying correlations, with different correlation structures favoring purely pairwise, purely hyperdegree, or mixed targeting approaches. II Configurational Higher-order network A system exhibiting higher-order interactions can be represented by a hypergraph H=(\mathcal{N},\mathcal{E}) , where \mathcal{N} is the set of N=|\mathcal{N}| nodes and \mathcal{E} is the set of E=|\mathcal{E}| hyperedges representing their interactions. Each hyperedge e\in\mathcal{E} is a subset of \mathcal{N} and can be characterized by its order m=|e|-1 . For example, m=1 corresponds to pairwise interactions, m=2 corresponds to group interactions of 3 nodes, and so on. The generalized hyperdegree vector of a node i is \mathbf{k}=(k_{1}^{i},k_{2}^{i},\ldots,k_{M}^{i}) , where k_{m}^{i} denotes the number of m -hyperedges incident to a node 12 . We define a joint probability distribution \mathbf{P}(\mathbf{k}) that gives the probability of a randomly chosen node having a hyperdegree vector \mathbf{k}=(k_{1},k_{2},\ldots,k_{M}) , where M is the maximum order considered. This joint distribution has known marginal distributions P_{m}(k_{m}) for each order of interactions m . A key aspect of

Read the complete text