Márton Pósfai
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Márton Pósfai has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Ivan Bonamassa
Department of Network and Data Science, Central European University, Vienna, Austria
3 shared publications - Gábor Pete
HUN-REN Alfréd Rényi Institute of Mathematics, Budapest, Hungary
2 shared publications - Luka Blagojević
Department of Network and Data Science, Central European University, Vienna, Austria
2 shared publications Sigurdur Örn Stefánsson
University of Iceland, Reykjavik, Iceland
2 shared publicationsÁdám Timár
HUN-REN Alfréd Rényi Institute of Mathematics, Budapest, Hungary
2 shared publications- Albert-László Barabási
Network Science Institute, Northeastern University, Boston, MA, USA.
1 shared publication
Latest Publications
- Open Access
- 15 August 2025
- Published by Springer Nature in Communications Physics
- Vol. 8 (1) , 1-13
Network dismantling by physical damage
It is well-understood that the network structure of complex systems affects their robustness; the role played by the shape of spatially embedded networks, however, is less explored. Here, we study the robustness of networks where links are physical objects or physically transfer some quantity, hence the links can be disrupted at any point along their trajectory. To model physical damage, we tile each network with boxes and we sequentially damage these boxes, removing any link from the network that intersects a damaged tile. Using model and empirical networks, we systematically explore how the layout and the structure of networks jointly affect the resulting percolation transition. For example, we analytically and numerically show that randomly damaging a vanishing fraction of tiles is enough to destroy large-scale connectivity in randomly embedded networks. This demonstrates that the presence of long-range links makes networks extremely vulnerable to physical damage. Our work contributes to the emergent theory of physical networks.
- 15 August 2025
- Published by Springer Nature in Communications Physics
- Vol. 8 (1) , 1-13
Open AccessNetwork dismantling by physical damage
It is well-understood that the network structure of complex systems affects their robustness; the role played by the shape of spatially embedded networks, however, is less explored. Here, we study the robustness of networks where links are physical objects or physically transfer some quantity, hence the links can be disrupted at any point along their trajectory. To model physical damage, we tile each network with boxes and we sequentially damage these boxes, removing any link from the network that intersects a damaged tile. Using model and empirical networks, we systematically explore how the layout and the structure of networks jointly affect the resulting percolation transition. For example, we analytically and numerically show that randomly damaging a vanishing fraction of tiles is enough to destroy large-scale connectivity in randomly embedded networks. This demonstrates that the presence of long-range links makes networks extremely vulnerable to physical damage. Our work contributes to the emergent theory of physical networks.
- Open Access
- 3 June 2025
- Published by American Physical Society (APS) in Physical Review Research
- Vol. 7 (2)
Random graphs with specific degree distribution and giant component size
Random networks are a powerful tool in the analytical modeling of complex networks as they allow us to write approximate mathematical models for diverse properties and behaviors of complex systems. These models are often used to study stochastic processes like percolation, where the giant connected component breaks down as edges are removed, yet they fail to properly account for the size of that component, even in a deterministic setting where all edges exist. Here, we introduce a simple conceptual step to account for such connectivity constraints in existing models. We distinguish between network neighbors based on two types of connections that can lead, or not, to the giant component, which we refer to as critical and subcritical degrees. The giant component is the largest unique component of a network that scales with the network's size under our model. It is analogous to many properties of interest, such as the largest epidemic possible on a contact network or the connectivity of an infrastructure network. Accounting explicitly for this component also allows us to capture important structural features of the network in a system of only one or two equations. When applied to sparse connected networks, we show that our approach compares favorably with the predictions of state-of-the-art models, like message passing, which require a number of equations that are linear in system size. We discuss potential applications of this simple framework for studying infrastructure networks where connectivity constraints are critical to the function of the system. Published by the American Physical Society 2025
- 3 June 2025
- Published by American Physical Society (APS) in Physical Review Research
- Vol. 7 (2)
Open AccessRandom graphs with specific degree distribution and giant component size
Random networks are a powerful tool in the analytical modeling of complex networks as they allow us to write approximate mathematical models for diverse properties and behaviors of complex systems. These models are often used to study stochastic processes like percolation, where the giant connected component breaks down as edges are removed, yet they fail to properly account for the size of that component, even in a deterministic setting where all edges exist. Here, we introduce a simple conceptual step to account for such connectivity constraints in existing models. We distinguish between network neighbors based on two types of connections that can lead, or not, to the giant component, which we refer to as critical and subcritical degrees. The giant component is the largest unique component of a network that scales with the network's size under our model. It is analogous to many properties of interest, such as the largest epidemic possible on a contact network or the connectivity of an infrastructure network. Accounting explicitly for this component also allows us to capture important structural features of the network in a system of only one or two equations. When applied to sparse connected networks, we show that our approach compares favorably with the predictions of state-of-the-art models, like message passing, which require a number of equations that are linear in system size. We discuss potential applications of this simple framework for studying infrastructure networks where connectivity constraints are critical to the function of the system. Published by the American Physical Society 2025
- Open Access
Network dismantling by physical damage
We explore the robustness of complex networks against physical damage. We focus on spatially embedded network models and datasets where links are physical objects or physically transfer some quantity, which can be disrupted at any point along its trajectory. To simulate physical damage, we tile the networks with boxes of equal size and sequentially damage them. By introducing an intersection graph to keep track of the links passing through tiles, we systematically analyze the connectivity of the network and explore how the physical layout and the topology of the network jointly affect its percolation threshold. We show that random layouts make networks extremely vulnerable to physical damage, driven by the presence of very elongated links, and that higher-dimensional embeddings further increase their vulnerability. We compare this picture against targeted physical damages, showing that it accelerates network dismantling and yields non-trivial geometric patterns. Finally, we apply our framework to several empirical networks, from airline networks to vascular systems and the brain, showing qualitative agreement with the theoretical predictions.
Open AccessNetwork dismantling by physical damage
We explore the robustness of complex networks against physical damage. We focus on spatially embedded network models and datasets where links are physical objects or physically transfer some quantity, which can be disrupted at any point along its trajectory. To simulate physical damage, we tile the networks with boxes of equal size and sequentially damage them. By introducing an intersection graph to keep track of the links passing through tiles, we systematically analyze the connectivity of the network and explore how the physical layout and the topology of the network jointly affect its percolation threshold. We show that random layouts make networks extremely vulnerable to physical damage, driven by the presence of very elongated links, and that higher-dimensional embeddings further increase their vulnerability. We compare this picture against targeted physical damages, showing that it accelerates network dismantling and yields non-trivial geometric patterns. Finally, we apply our framework to several empirical networks, from airline networks to vascular systems and the brain, showing qualitative agreement with the theoretical predictions.
- Open Access
- 19 September 2024
- Published by Springer Nature in Nature Communications
- Vol. 15 (1) , 1
Author Correction: Physical networks as network-of-networks
- 19 September 2024
- Published by Springer Nature in Nature Communications
- Vol. 15 (1) , 1
Open AccessAuthor Correction: Physical networks as network-of-networks
- Open Access
- 7 June 2024
- Published by Springer Nature in Nature Communications
- Vol. 15 (1) , 1-8
Physical networks as network-of-networks
Physical networks are made of nodes and links that are physical objects embedded in a geometric space. Understanding how the mutual volume exclusion between these elements affects the structure and function of physical networks calls for a suitable generalization of network theory. Here, we introduce a network-of-networks framework where we describe the shape of each extended physical node as a network embedded in space and these networks are bound together by physical links. Relying on this representation, we introduce a minimal model of network growth and we show for a general class of physical networks that volume exclusion induces heterogeneity in both node volume and degree, with the two becoming correlated. These emergent properties strongly affect the dynamics on physical networks: by calculating their Laplacian spectrum as a function of the coupling strength between the nodes we show that degree-volume correlations suppress the role of hubs as early spreaders in diffusive dynamics. We apply the network-of-networks framework to describe several real systems and find properties analog to the minimal model networks. The prevalence of these properties points towards general growth mechanisms that do not depend on the specifics of the systems.
- 7 June 2024
- Published by Springer Nature in Nature Communications
- Vol. 15 (1) , 1-8
Open AccessPhysical networks as network-of-networks
Physical networks are made of nodes and links that are physical objects embedded in a geometric space. Understanding how the mutual volume exclusion between these elements affects the structure and function of physical networks calls for a suitable generalization of network theory. Here, we introduce a network-of-networks framework where we describe the shape of each extended physical node as a network embedded in space and these networks are bound together by physical links. Relying on this representation, we introduce a minimal model of network growth and we show for a general class of physical networks that volume exclusion induces heterogeneity in both node volume and degree, with the two becoming correlated. These emergent properties strongly affect the dynamics on physical networks: by calculating their Laplacian spectrum as a function of the coupling strength between the nodes we show that degree-volume correlations suppress the role of hubs as early spreaders in diffusive dynamics. We apply the network-of-networks framework to describe several real systems and find properties analog to the minimal model networks. The prevalence of these properties points towards general growth mechanisms that do not depend on the specifics of the systems.
- Open Access
- 17 January 2024
- Published by The Royal Society in Royal Society Open Science
- Vol. 11 (1) , 231460
What you have, not who you know: food-enhanced social capital and changes in social behavioural relationships in a non-human primate
Social network position in non-human primates has far-reaching fitness consequences. Critically, social networks are both heterogeneous and dynamic, meaning an individual's current network position is likely to change due to both intrinsic and extrinsic factors. However, our understanding of the drivers of changes in social network position is largely confined to opportunistic studies. Experimental research on the consequences of in situ, controlled network perturbations is limited. Here we conducted a food-based experiment in rhesus macaques to assess whether allowing an individual the ability to provide high-quality food to her group changed her social behavioural relationships. We considered both her social network position across five behavioural networks, as well as her dominance and kin interactions. We found that gaining control over a preferential food resource had far-reaching social consequences. There was an increase in both submission and aggression centrality and changes in the socio-demographic characteristics of her agonistic interaction partners. Further, we found that her grooming balance shifted in her favour as she received more grooming than she gave. Together, these results provide a novel, preliminary insight into how in situ , experimental manipulations can modify social network position and point to broader network-level shifts in both social capital and social power.
- 17 January 2024
- Published by The Royal Society in Royal Society Open Science
- Vol. 11 (1) , 231460
Open AccessWhat you have, not who you know: food-enhanced social capital and changes in social behavioural relationships in a non-human primate
Social network position in non-human primates has far-reaching fitness consequences. Critically, social networks are both heterogeneous and dynamic, meaning an individual's current network position is likely to change due to both intrinsic and extrinsic factors. However, our understanding of the drivers of changes in social network position is largely confined to opportunistic studies. Experimental research on the consequences of in situ, controlled network perturbations is limited. Here we conducted a food-based experiment in rhesus macaques to assess whether allowing an individual the ability to provide high-quality food to her group changed her social behavioural relationships. We considered both her social network position across five behavioural networks, as well as her dominance and kin interactions. We found that gaining control over a preferential food resource had far-reaching social consequences. There was an increase in both submission and aggression centrality and changes in the socio-demographic characteristics of her agonistic interaction partners. Further, we found that her grooming balance shifted in her favour as she received more grooming than she gave. Together, these results provide a novel, preliminary insight into how in situ , experimental manipulations can modify social network position and point to broader network-level shifts in both social capital and social power.
- Top Cited Papers
- 6 November 2023
- Published by Springer Nature in Nature Physics
- Vol. 20 (1) , 142-149
Impact of physicality on network structure
- 6 November 2023
- Published by Springer Nature in Nature Physics
- Vol. 20 (1) , 142-149
Top Cited PapersImpact of physicality on network structure