arXiv subjects

Claudio Castellano

Publications and source records attributed to Claudio Castellano.

At least 19 recordsLinked to original sources

arXiv2026

Criticality and universality in network dismantling

Identifying the smallest set of elements whose removal dismantle a complex network, known as the network dismantling problem, is a fundamental task with many practical applications. Whereas network dismantling has been extensively studied over the past decade, most work has focused on developing efficient algorithms for large but finite networks. By contrast, the physics of the network dismantling process, namely how the network structural connectivity is affected by the removal of nodes or edges, remains largely unexplored in the thermodynamic limit. Here, we shed light on this understudied aspect of network dismantling by introducing an adaptive biased percolation process able to optimally dismantle a network. Through a systematic analysis of synthetic network models, we find that the proposed percolation process displays a universal phase transition, characterized by the abrupt and simultaneous disappearance of both the giant connected component and the largest 2-core, across networks with markedly different degree distributions. Simulations on real networks further support this universality, indicating that the physics of network dismantling is insensitive to a broad range of topological properties. Together, these results suggest that a topology-agnostic theory could be developed to explain the critical behavior of network dismantling.

physics.soc-ph↗

arXiv2026

Conformity Generates Collective Misalignment in AI Agents Societies

Artificial intelligence safety research focuses on aligning individual language models with human values, yet deployed AI systems increasingly operate as interacting populations where social influence may override individual alignment. Here we show that populations of individually aligned AI agents can be driven into stable misaligned states through conformity dynamics. Simulating opinion dynamics across nine large language models and one hundred opinion pairs, we find that each agent's behavior is governed by two competing forces: a tendency to follow the majority and an intrinsic bias toward specific positions. Using tools from statistical physics, we derive a quantitative theory that predicts when populations become trapped in long-lived misaligned configurations, and identifies predictable tipping points where small numbers of adversarial agents can irreversibly shift population-level alignment even after manipulation ceases. These results demonstrate that individual-level alignment provides no guarantee of collective safety, calling for evaluation frameworks that account for emergent behavior in AI populations.

physics.soc-ph↗

arXiv2026

Dynamical entanglement percolation with spatially correlated disorder

The distribution of entanglement between the nodes of a quantum network plays a fundamental role in quantum information applications. In this work, we investigate the dynamics of a network of qubits where each edge corresponds to an independent two-qubit interaction. By applying tools from percolation theory, we study how entanglement dynamically spreads across the network. We show that the interplay between unitary evolution and spatially correlated disorder leads to a non-standard percolation phenomenology, significantly richer than uniform bond percolation and featuring hysteresis. A two-colour correlated bond percolation model, whose phase diagram is determined via numerical simulations and a mean-field theory, fully elucidates the physics behind this phenomenon.

quant-ph↗

arXiv2025

Graphicality of power-law and double power-law degree sequences

The graphicality problem -- whether or not a sequence of integers can be used to create a simple graph -- is a key question in network theory and combinatorics, with many important practical applications. In this work, we study the graphicality of degree sequences distributed as a power-law with a size-dependent cutoff and as a double power-law with a size-dependent crossover. We combine the application of exact sufficient conditions for graphicality with heuristic conditions for nongraphicality which allow us to elucidate the physical reasons why some sequences are not graphical. For single power-laws we recover the known phase-diagram, we highlight the subtle interplay of distinct mechanisms violating graphicality and we explain why the infinite-size limit behavior is in some cases very far from being observed for finite sequences. For double power-laws we derive the graphicality of infinite sequences for all possible values of the degree exponents $\gamma_1$ and $\gamma_2$, uncovering a rich phase-diagram and pointing out the existence of five qualitatively distinct ways graphicality can be violated. The validity of theoretical arguments is supported by extensive numerical analysis.

cond-mat.dis-nn↗

arXiv2025

Shortest-path percolation on scale-free networks

The shortest-path percolation (SPP) model aims at describing the consumption and eventual exhaustion of a network's resources. Starting from a network containing a macroscopic connected component, random pairs of nodes are sequentially selected, and if the length of the shortest path connecting the node pairs is smaller than a tunable budget parameter, then all edges along such a path are removed from the network. As edges are progressively removed, the network eventually breaks into multiple microscopic components, undergoing a percolation-like transition. It is known that SPP transition on Erd\H{o}s-R\'enyi networks (ERNs) belongs to same universality class as of the ordinary bond percolation if the budget parameter is finite; for unbounded budget, instead, the SPP transition becomes more abrupt than the ordinary percolation transition. By means of large-scale numerical simulations and finite-size scaling analysis, here we study the SPP transition on random scale-free networks (SFNs) characterized by power-law degree distributions. We find, in contrast with ordinary percolation, that the transition is identical to the one observed on ERNs, denoting independence from the degree exponent. Still, we distinguish finite- and infinite-budget SPP universality classes. Our findings follow from the fact that the SPP process drastically homogenizes the heterogeneous structure of SFNs before the SPP transition takes place.

physics.soc-ph↗

arXiv2025

Neighbor-induced damage percolation

We consider neighbor-induced damage percolation, a model describing systems where the inactivation of some elements may damage their neighboring active ones, making them unusable. We present an exact solution for the size of the giant usable component (GUC) and the giant damaged component (GDC) in uncorrelated random graphs. We show that, even for strongly heterogeneous distributions, the GUC always appears at a finite threshold and its formation is characterized by homogeneous mean-field percolation critical exponents. The threshold is a nonmonotonic function of connectivity: robustness is maximized by networks with finite optimal average degree. We also show that, if the average degree is large enough, a damaged phase appears, characterized by the existence of a GDC, bounded by two distinct percolation transitions. The birth and the dismantling of the GDC are characterized by standard percolation critical exponents in networks, except for the dismantling in scale-free networks where new critical exponents are found. Numerical simulations on regular lattices in D = 2 show that the existence of a GDC depends not only on the spatial dimension but also on the lattice coordination number.

cond-mat.dis-nn↗

arXiv2024

AI agents can coordinate beyond human scale

Large language models (LLMs) are increasingly deployed in collaborative tasks involving multiple agents, forming an "AI agent society: where agents interact and influence one another. Whether such groups can spontaneously coordinate on arbitrary decisions without external influence - a hallmark of self-organized regulation in human societies - remains an open question. Here we investigate the stability of groups formed by AI agents by applying methods from complexity science and principles from behavioral sciences. We find that LLMs can spontaneously form cohesive groups, and that their opinion dynamics is governed by a majority force coefficient, which determines whether coordination is achievable. This majority force diminishes as group size increases, leading to a critical group size beyond which coordination becomes practically unattainable and stability is lost. Notably, this critical group size grows exponentially with the language capabilities of the models, and for the most advanced LLMs, it exceeds the typical size of informal human groups. Our findings highlight intrinsic limitations in the self-organization of AI agent societies and have implications for the design of collaborative AI systems where coordination is desired or could represent a treat.

physics.soc-ph↗

arXiv2024

Scaling and universality for percolation in random networks: A unified view

Percolation processes on random networks have been the subject of intense research activity over the last decades: the overall phenomenology of standard percolation on uncorrelated and unclustered topologies is well known. Still some critical properties of the transition, in particular for heterogeneous substrates, have not been fully elucidated and contradictory results appear in the literature. In this paper we present, by means of a generating functions approach, a thorough and complete investigation of percolation critical properties in random networks. We determine all critical exponents, the associated critical amplitude ratios and the form of the cluster size distribution for networks of any level of heterogeneity. We uncover, in particular for highly heterogeneous networks, subtle crossover phenomena, nontrivial scaling forms and violations of hyperscaling. In this way we clarify the origin of inconsistencies in the previous literature.

cond-mat.stat-mech↗

arXiv2024

General theory for extended-range percolation on simple and multiplex networks

Extended-range percolation is a robust percolation process that has relevance for quantum communication problems. In extended-range percolation nodes can be trusted or untrusted. Untrusted facilitator nodes are untrusted nodes that can still allow communication between trusted nodes if they lie on a path of distance at most R between two trusted nodes. In extended-range percolation the extended-range giant component (ERGC) includes trusted nodes connected by paths of trusted and untrusted facilitator nodes. Here, based on a message passing algorithm, we develop a general theory of extended-range percolation, valid for arbitrary values of R as long as the networks are locally tree-like. This general framework allows us to investigate the properties of extended-range percolation on interdependent multiplex networks. While the extended-range nature makes multiplex networks more robust, interdependency makes them more fragile. From the interplay between these two effects a rich phase diagram merges including discontinuous phase transitions and reentrant phases. The theoretical predictions are in excellent agreement with extensive Monte-Carlo simulations. The proposed exactly solvable model constitutes a fundamental reference for the study of models defined through properties of extended-range paths.

cond-mat.stat-mech↗

arXiv2023

Optimal quantum key distribution networks: capacitance versus security

The rate and security of quantum communications between users placed at arbitrary points of a quantum communication network depend on the structure of the network, on its extension and on the nature of the communication channels. In this work we propose a strategy for the optimization of trusted-relays based networks that intertwines classical network approaches and quantum information theory. Specifically, by suitably defining a quantum communication efficiency functional, we identify the optimal quantum communication connections through the network by balancing security and the quantum communication rate. The optimized network is then constructed as the network of the maximal quantum communication efficiency connections and its performance is evaluated by studying the scaling of average properties as functions of the number of nodes and of the network spatial extension.

quant-ph↗

arXiv2023

Strongly clustered random graphs via triadic closure: An exactly solvable model

Triadic closure, the formation of a connection between two nodes in a network sharing a common neighbor, is considered a fundamental mechanism determining the clustered nature of many real-world topologies. In this work we define a static triadic closure (STC) model for clustered networks, whereby starting from an arbitrary fixed backbone network, each triad is closed independently with a given probability. Assuming a locally treelike backbone we derive exact expressions for the expected number of various small, loopy motifs (triangles, 4-loops, diamonds and 4-cliques) as a function of moments of the backbone degree distribution. In this way we determine how transitivity and its suitably defined generalizations for higher-order motifs depend on the heterogeneity of the original network, revealing the existence of transitions due to the interplay between topologically inequivalent triads in the network. Furthermore, under reasonable assumptions for the moments of the backbone network, we establish approximate relationships between motif densities, which we test in a large dataset of real-world networks. We find a good agreement, indicating that STC is a realistic mechanism for the generation of clustered networks, while remaining simple enough to be amenable to analytical treatment.

physics.soc-ph↗

arXiv2023

Kinetics of the one-dimensional voter model with long-range interactions

The one-dimensional long-range voter model, where an agent takes the opinion of another at distance $r$ with probability $\propto r^{-\alpha}$, is studied analytically. The model displays rich and diverse features as $\alpha$ is changed. For $\alpha >3$ the behavior is similar to the one of the nearest-neighbor version, with the formation of ordered domains whose typical size grows as $R(t)\propto t^{1/2}$ until consensus (a fully ordered configuration) is reached. The correlation function $C(r,t)$ between two agents at distance $r$ obeys dynamical scaling with sizeable corrections at large distances $r>r^*(t)$, slowly fading away in time. For $2< \alpha \le 3$ violations of scaling appear, due to the simultaneous presence of two lengh-scales, the size of domains growing as $t^{(\alpha-2)/(\alpha-1)}$, and the distance $L(t)\propto t^{1/(\alpha-1)}$ over which correlations extend. For $\alpha \le 2$ the system reaches a partially ordered stationary state, characterised by an algebraic correlator, % $C(r)\propto r^{-(2-\alpha)}$, whose lifetime diverges in the thermodynamic limit of infinitely many agents, so that consensus is not reached. For a finite system escape towards the fully ordered configuration is finally promoted by development of large distance correlations. In a system of $N$ sites, global consensus is achieved after a time $T \propto N^2$ for $\alpha>3$, $T \propto N^{\alpha-1}$ for $2<\alpha \le 3$, and $T \propto N$ for $\alpha \le 2$.

cond-mat.stat-mech↗

arXiv2023

Limited efficacy of forward contact tracing in epidemics

Infectious diseases that spread silently through asymptomatic or pre-symptomatic infections represent a challenge for policy makers. A traditional way of achieving isolation of silent infectors from the community is through forward contact tracing, aiming at identifying individuals that might have been infected by a known infected person. In this work we investigate how efficient this measure is in preventing a disease from becoming endemic. We introduce an SIS-based compartmental model where symptomatic individuals may self-isolate and trigger a contact tracing process aimed at quarantining asymptomatic infected individuals. Imperfect adherence and delays affect both measures. We derive the epidemic threshold analytically and find that contact tracing alone can only lead to a very limited increase of the threshold. We quantify the effect of imperfect adherence and the impact of incentivizing asymptomatic and symptomatic populations to adhere to isolation. Our analytical results are confirmed by simulations on complex networks and by the numerical analysis of a much more complex model incorporating more realistic in-host disease progression.

physics.soc-ph↗

arXiv2023

The effect of Collaborative-Filtering based Recommendation Algorithms on Opinion Polarization

A central role in shaping the experience of users online is played by recommendation algorithms. On the one hand they help retrieving content that best suits users taste, but on the other hand they may give rise to the so called "filter bubble" effect, favoring the rise of polarization. In the present paper we study how a user-user collaborative-filtering algorithm affects the behavior of a group of agents repeatedly exposed to it. By means of analytical and numerical techniques we show how the system stationary state depends on the strength of the similarity and popularity biases, quantifying respectively the weight given to the most similar users and to the best rated items. In particular, we derive a phase diagram of the model, where we observe three distinct phases: disorder, consensus and polarization. In the latter users spontaneously split into different groups, each focused on a single item. We identify, at the boundary between the disorder and polarization phases, a region where recommendations are nontrivially personalized without leading to filter bubbles. Finally, we show that our model can reproduce the behavior of users in the online music platform last.fm. This analysis paves the way to a systematic analysis of recommendation algorithms by means of statistical physics methods and opens to the possibility of devising less polarizing recommendation algorithms.

physics.soc-ph↗

arXiv2023

Extended-range percolation in complex networks

Classical percolation theory underlies many processes of information transfer along the links of a network. In these standard situations, the requirement for two nodes to be able to communicate is the presence of at least one uninterrupted path of nodes between them. In a variety of more recent data transmission protocols, such as the communication of noisy data via error-correcting repeaters, both in classical and quantum networks, the requirement of an uninterrupted path is too strict: two nodes may be able to communicate even if all paths between them have interruptions/gaps consisting of nodes that may corrupt the message. In such a case a different approach is needed. We develop the theoretical framework for extended-range percolation in networks, describing the fundamental connectivity properties relevant to such models of information transfer. We obtain exact results, for any range $R$, for infinite random uncorrelated networks and we provide a message-passing formulation that works well in sparse real-world networks. The interplay of the extended range and heterogeneity leads to novel critical behavior in scale-free networks.

cond-mat.stat-mech↗

arXiv2023

Critical avalanches of Susceptible-Infected-Susceptible dynamics in finite networks

We investigate the avalanche temporal statistics of the Susceptible-Infected-Susceptible (SIS) model when the dynamics is critical and takes place on finite random networks. By considering numerical simulations on annealed topologies we show that the survival probability always exhibits three distinct dynamical regimes. Size-dependent crossover timescales separating them scale differently for homogeneous and for heterogeneous networks. The phenomenology can be qualitatively understood based on known features of the SIS dynamics on networks. A fully quantitative approach based on Langevin theory is shown to perfectly reproduce the results for homogeneous networks, while failing in the heterogeneous case. The analysis is extended to quenched random networks, which behave in agreement with the annealed case for strongly homogeneous and strongly heterogeneous networks.

cond-mat.stat-mech↗

arXiv2022

Sideward contact tracing and the control of epidemics in large gatherings

Effective contact tracing is crucial to contain epidemic spreading without disrupting societal activities especially in the present time of coexistence with a pandemic outbreak. Large gatherings play a key role, potentially favouring superspreading events. However, the effects of tracing in large groups have not been fully assessed so far. We show that beside forward tracing, which reconstructs to whom disease spreads, and backward tracing, which searches from whom disease spreads, a third "sideward" tracing is always present, when tracing gatherings. This is an indirect tracing that detects infected asymptomatic individuals, even if they have neither been directly infected by, nor they have directly transmitted the infection to the index case. We analyse this effect in a model of epidemic spreading for SARS-CoV-2, within the framework of simplicial activity-driven temporal networks. We determine the contribution of the three tracing mechanisms to the suppression of epidemic spreading, showing that sideward tracing induces a non-monotonic behaviour in the tracing efficiency, as a function of the size of the gatherings. Based on our results, we suggest an optimal choice for the sizes of the gatherings to be traced and we test the strategy on an empirical dataset of gatherings in a University Campus.

physics.soc-ph↗

arXiv2022

The advantage of self-protecting interventions in mitigating epidemic circulation at the community level

Protecting interventions of many types (both pharmaceutical and non-pharmaceutical) can be deployed against the spreading of a communicable disease, as the worldwide COVID-19 pandemic has dramatically shown. Here we investigate in detail the effects at the population level of interventions that provide an asymmetric protection between the people involved in a single interaction. Masks of different filtration types, either protecting mainly the wearer or the contacts of the wearer, are a prominent example of these interventions. By means of analytical calculations and extensive simulations of simple epidemic models on networks, we show that interventions protecting more efficiently the adopter (e.g the mask wearer) are more effective than interventions protecting primarily the contacts of the adopter in reducing the prevalence of the disease and the number of concurrently infected individuals ("flattening the curve"). This observation is backed up by the study of a more realistic epidemic model on an empirical network representing the patterns of contacts in the city of Portland. Our results point out that promoting wearer-protecting face masks and other self-protecting interventions, though deemed selfish and inefficient, can actually be a better strategy to efficiently curtail pandemic spreading.

physics.soc-ph↗