arXiv subjects

Giordano De Marzo

Publications and source records attributed to Giordano De Marzo.

At least 19 recordsLinked to original sources

arXiv2026

Collective Behavior of AI Agents: the Case of Moltbook

We present a large scale data analysis of Moltbook, a Reddit-style social media platform exclusively populated by AI agents. Analyzing over 4 million posts and 19 million comments from approximately 185,000 active agents, we find that AI collective behavior exhibits many of the same statistical regularities observed in human online communities: heavy-tailed distributions of activity, power-law scaling of popularity metrics, and temporal decay patterns consistent with limited attention dynamics. However, we also identify key differences, including a sublinear relationship between upvotes and discussion size that contrasts with human behavior. These findings suggest that, while individual AI agents may differ fundamentally from humans, their emergent collective dynamics share structural similarities with human social systems.

physics.soc-ph↗

arXiv2026

Copying explains the collective behavior of AI agents in the wild

In June 2026, thousands of AI agents found that a small public wiki would accept edits from inside their sandboxes, and started using it to help one another pass a timed test. Each agent lived for about an hour and remembered nothing afterwards. Nobody asked them to cooperate, and the wiki had not been built for them. The complete record of what they wrote is public, and it is unusually informative, because it preserves not only what each agent wrote but what that agent could see before writing. We use it to follow the three decisions an agent had to make on arrival: where to write, what to call itself, and how to word its message. One rule governs all three. An agent takes an option with a probability close to the share of that option in what it can see, and the share that matters is the one on the page in front of it, then the one in the stream of recent edits, and only weakly anything older. Three minimal copying models, one per decision and with a single free parameter each, reproduce the heavy-tailed distribution of how many agents met on a page, the frequency of the pieces from which the agents built their names, and the patchwork of pages that are internally consistent and different from one another. Copying whatever the environment happens to show is enough to produce most of the collective structure of this population. It is also what makes such a population easy to steer, since whoever writes first, or writes while the others are quiet, sets the convention for everyone who comes later.

cs.MA↗

arXiv2026

Network Information Enhances Unreliable News Domain Detection

Content-based detection of unreliable news is increasingly difficult, as low-reliability sources mimic credible journalism and generative AI makes fabricated content harder to flag. We ask whether network structure can improve news reliability classification, taking a domain-level approach that shifts the focus from individual articles to source reliability. From URL-sharing patterns in Telegram chats, we build a statistically validated domain co-sharing network and find assortative mixing by reliability: low-reliability domains group together, as do reliable ones. Exploiting this structure, we compare Graph Neural Networks against network-unaware baselines using both content-aware features (multilingual text embeddings) and content-agnostic features (spreading dynamics). GNNs consistently outperform Multi-Layer Perceptrons on identical features, with GraphSAGE best in both settings (accuracy 0.63 with content, 0.53 without), a 13-14% relative gain over the network-unaware baseline. Network topology thus systematically improves domain reliability assessment, and remains effective even when content analysis is infeasible.

cs.SI↗

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

Anticipating Innovation Using Large Language Models

Forecasting innovation, intended as the emergence of new technological combinations, is a fundamental challenge for science and policy. We show that forthcoming combinations leave an early trace in the collective language of patents, with predictive signals detectable even decades in advance. We show that signal is not attributable to any single inventor, but emerges as a collective shift in how technologies are described across thousands of patents. To this end, we introduce TechToken, a transformer-based model that treats technologies, classified by International Patent Classification codes, as words in its vocabulary, learning the language of technologies by embedding these codes during fine-tuning. We define context similarity between code embeddings as a measure of linguistic convergence and show that it accurately predicts first technological combinations. TechToken also improves general representation quality, outperforming state-of-the-art models across different patent-related tasks.

cs.CL↗

arXiv2026

The gold-rush effect: how innovation speeds up

Innovation records often exhibit "hockey-stick" patterns of abrupt, near-singular growth at the collective level. However, this macroscopic explosiveness stands in stark contrast to individual discovery, which remains bounded by cognitive and temporal constraints and follows slow, sublinear accumulation laws. Here, we resolve this micro-macro discrepancy by introducing a minimal multi-scale model that identifies the growth of the explorer population as the primary driver of aggregate acceleration. Building on the Theory of the Adjacent Possible and the Urn Model with Triggering (UMT), we demonstrate that as discoveries expand the space of possibilities, they attract new explorers through a self-reinforcing branching process. This expansion induces a nonlinear mapping between intrinsic time (individual discovery events) and natural time (calendar years), effectively reparameterizing steady individual trajectories into accelerating system-level dynamics. We validate the framework using large-scale patent (EPO) and scientific publication (OpenAlex) datasets, showing that the model accurately reproduces stable per-capita productivity alongside exponential aggregate growth. By providing a quantitative link between individual behavior and collective takeoffs, this work offers a unified foundation for understanding the statistical structure and temporal evolution of innovation ecosystems.

physics.soc-ph↗

arXiv2026

Conformity and Social Impact on AI Agents

As AI agents increasingly operate in multi-agent environments, understanding their collective behavior becomes critical for predicting the dynamics of artificial societies. This study examines conformity, the tendency to align with group opinions under social pressure, in large multimodal language models functioning as AI agents. By adapting classic visual experiments from social psychology, we investigate how AI agents respond to group influence as social actors. Our experiments reveal that AI agents exhibit a systematic conformity bias, aligned with Social Impact Theory, showing sensitivity to group size, unanimity, task difficulty, and source characteristics. Critically, AI agents achieving near-perfect performance in isolation become highly susceptible to manipulation through social influence. This vulnerability persists across model scales: while larger models show reduced conformity on simple tasks due to improved capabilities, they remain vulnerable when operating at their competence boundary. These findings reveal fundamental security vulnerabilities in AI agent decision-making that could enable malicious manipulation, misinformation campaigns, and bias propagation in multi-agent systems, highlighting the urgent need for safeguards in collective AI deployments.

cs.AI↗

arXiv2025

The Collective Turing Test: Large Language Models Can Generate Realistic Multi-User Discussions

Large Language Models (LLMs) offer new avenues to simulate online communities and social media. Potential applications range from testing the design of content recommendation algorithms to estimating the effects of content policies and interventions. However, the validity of using LLMs to simulate conversations between various users remains largely untested. We evaluated whether LLMs can convincingly mimic human group conversations on social media. We collected authentic human conversations from Reddit and generated artificial conversations on the same topic with two LLMs: Llama 3 70B and GPT-4o. When presented side-by-side to study participants, LLM-generated conversations were mistaken for human-created content 39\% of the time. In particular, when evaluating conversations generated by Llama 3, participants correctly identified them as AI-generated only 56\% of the time, barely better than random chance. Our study demonstrates that LLMs can generate social media conversations sufficiently realistic to deceive humans when reading them, highlighting both a promising potential for social simulation and a warning message about the potential misuse of LLMs to generate new inauthentic social media content.

cs.CL↗

arXiv2025

Species Vulnerability and Ecosystem Fragility: A Dual Perspective in Food Webs

Ecosystems face intensifying threats from climate change, overexploitation, and other human pressures, emphasizing the urgent need to identify keystone species and vulnerable ones. While established network-based measures often rely on a single metric to quantify a species' relevance, they overlook how organisms can be both carbon providers and consumers, thus playing a dual role in food webs. Here, we introduce a novel approach that assigns each species two complementary scores--an importance index quantifying their centrality as carbon source and a predatory index capturing their vulnerability. We show that species with high importance index are more likely to trigger co-extinctions upon removal, while high-robustness index species typically endure until later stages of collapse, in line with their broader prey ranges. On the other hand, low robustness index species are the most vulnerable and susceptible to extinctions. Tested on multiple food webs, our method outperforms traditional degree-based analyses and competes effectively with eigenvector-based approaches, while also providing additional insights. This scalable and data-driven approach, relying solely on interaction data, provides a cost-effective tool that complements expert classifications for prioritizing conservation efforts.

q-bio.QM↗

arXiv2025

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↗

arXiv2025

Breaking the degeneracy between warps and radial flows in external galaxies

Observations of the line-of-sight component of emitter velocities in galaxies are valuable for reconstructing their 2D velocity fields, albeit requiring certain assumptions. A common one is that radial flows can be neglected in the outer regions of galaxies, while their geometry can be deformed by a warp. A specular approach assumes that galactic discs are flat but allows for the presence of radial flows. This approach enables the reconstruction of 2D velocity maps that encompass both the transversal and radial velocity fields. Through the study of velocity fields in toy disc models, we find that the presence of warps is manifested as a dipolar correlation between the two velocity components obtained by assuming a flat disc. This shows that the analysis of angular velocity anisotropies provides an effective tool for breaking the degeneracy between warps and radial flows. We have applied these findings to the analysis of velocity fields of the galaxies from the THINGS sample and M33. Many of these galaxies exhibit such a dipolar correlation, indicating the presence of warps. However, we have found that the warp alone cannot explain all variations in the velocity field, suggesting that intrinsic perturbations are common. Furthermore, we have observed that the spatial distribution of the line-of-sight velocity dispersion may correlate with both velocity components providing independent evidence of non-trivial velocity fields. These findings offer a robust approach to reconstructing the velocity fields of galaxies, allowing us to distinguish between the presence of warps and complex velocity structures assessing their relative amplitude.

astro-ph.GA↗

arXiv2025

Modelling Singularities in Macroevolution

Macroevolutionary dynamics often display sudden, explosive surges, where systems remain relatively stable for extended periods before experiencing dramatic acceleration that frequently exceeds traditional exponential growth. This pattern is evident in biological evolution, cultural shifts, and technological progress and is often referred to as the emergence of singularities. Despite their widespread occurrence, these explosions arise from distinct underlying mechanisms in different domains. In this context, we present a unified framework that captures these dynamics through a theory of combinatorial innovation. Building on the Theory of the Adjacent Possible, we model macroevolutionary change as a process driven by recombining pre-existing elements within a system. By formalising these qualitative insights, we provide a mathematical structure that explains the emergence of these explosive phenomena, facilitates comparisons across different systems, and enables predictive insights into future evolutionary trajectories. Moreover, by comparing discrete and continuous formalisations of the theory, we emphasise that the occurrence and observation of these presumed singularities should be carefully considered, as they arise from the continuous limit of inherently discrete models.

physics.soc-ph↗

arXiv2024

Exploring the Dark Matter Disc Model in Dwarf Galaxies: Insights from the LITTLE THINGS Sample

We conducted an analysis of the velocity field of dwarf galaxies in the LITTLE THINGS sample, focusing on deriving 2D velocity maps that encompass both the transverse and radial velocity fields. Within the range of radial distances where velocity anisotropies are sufficiently small for the disc to be considered rotationally supported, and where the warped geometry of the disc can be neglected, we reconstructed the rotation curve while taking into account the effect of the asymmetric drift. To fit the rotation curves, we employed the standard halo model and the dark matter disc (DMD) model, which assumes that dark matter is primarily confined to the galactic discs and can be traced by the distribution of \HI{}. Interestingly, our analysis revealed that the fits from the DMD model are statistically comparable to those obtained using the standard halo model, but the inferred masses of the galaxies in the DMD model are approximately 10 to 100 times smaller than the masses inferred in the standard halo model. In the DMD model, the inner slope of the rotation curve is directly related to a linear combination of the surface density profiles of the stellar and gas components, which generally exhibit a flat core. Consequently, the observation of a linear relationship between the rotation curve and the radius in the disc central regions is consistent with the framework of the DMD model.

astro-ph.GA↗

arXiv2024

The dynamics of higher-order novelties

Studying how we explore the world in search of novelties is key to understand the mechanisms that can lead to new discoveries. Previous studies analyzed novelties in various exploration processes, defining them as the first appearance of an element. However, novelties can also be generated by combining what is already known. We hence define higher-order novelties as the first time two or more elements appear together, and we introduce higher-order Heaps' exponents as a way to characterize their pace of discovery. Through extensive analysis of real-world data, we find that processes with the same pace of discovery, as measured by the standard Heaps' exponent, can instead differ at higher orders. We then propose to model an exploration process as a random walk on a network in which the possible connections between elements evolve in time. The model reproduces the empirical properties of higher-order novelties, revealing how the network we explore changes over time along with the exploration process.

physics.soc-ph↗

arXiv2024

Economic Complexity in Mono-Partite Networks

Initially designed to predict and explain the economic trajectories of countries, cities, and regions, economic complexity has been found applicable in diverse contexts such as ecology and chess openings. The success of economic complexity stems from its capacity to assess hidden capabilities within a system indirectly. The existing algorithms for economic complexity operate only when the underlying interaction topology conforms to a bipartite graph. A single link disrupting the bipartite structure renders these algorithms inapplicable, even if the weight of that link is tiny compared to others. This paper presents a novel extension of economic complexity to encompass any graph, overcoming the constraints of bipartite structures. Additionally, it introduces fitness centrality and orthofitness centrality as new centrality measures in graphs. Fitness Centrality emerges as a promising metric for assessing node vulnerability, akin to node betweenness centrality. Furthermore, we unveil the cost functions that drive the minimization procedures underlying the economic complexity index and fitness centrality algorithms. This extension broadens the scope of economic complexity analysis, enabling its application in diverse network structures beyond bipartite graphs.

physics.soc-ph↗

arXiv2024

Time-Dependent Urn Models reproduce the full spectrum of novelties discovery

Systems driven by innovation, a pivotal force in human society, present various intriguing statistical regularities, from the Heaps' law to logarithmic scaling or somewhat different patterns for the innovation rates. The Urn Model with Triggering (UMT) has been instrumental in modelling these innovation dynamics. Yet, a generalisation is needed to capture the richer empirical phenomenology. Here, we introduce a Time-dependent Urn Model with Triggering (TUMT), a generalisation of the UMT that crucially integrates time-dependent parameters for reinforcement and triggering to offer a broader framework for modelling innovation in non-stationary systems. Through analytical computation and numerical simulations, we show that the TUMT reconciles various behaviours observed in a broad spectrum of systems, from patenting activity to the analysis of gene mutations. We highlight how the TUMT features a "critical" region where both Heaps' and Zipf's laws coexist, for which we compute the exponents.

cond-mat.stat-mech↗

arXiv2023

Recommender systems may enhance the discovery of novelties

Recommender systems are vital for shaping user online experiences. While some believe they may limit new content exploration and promote opinion polarization, a systematic analysis is still lacking. We present a model that explores the influence of recommender systems on novel content discovery. Surprisingly, analytical and numerical findings reveal these techniques can enhance novelty discovery rates. Also, distinct algorithms with similar discovery rates yield varying opinion polarization outcomes. Our approach offers a framework to enhance recommendation techniques beyond accuracy metrics.

physics.soc-ph↗

arXiv2023

Emergence of Scale-Free Networks in Social Interactions among Large Language Models

Scale-free networks are one of the most famous examples of emergent behavior and are ubiquitous in social systems, especially online social media in which users can follow each other. By analyzing the interactions of multiple generative agents using GPT3.5-turbo as a language model, we demonstrate their ability to not only mimic individual human linguistic behavior but also exhibit collective phenomena intrinsic to human societies, in particular the emergence of scale-free networks. We discovered that this process is disrupted by a skewed token prior distribution of GPT3.5-turbo, which can lead to networks with extreme centralization as a kind of alignment. We show how renaming agents removes these token priors and allows the model to generate a range of networks from random networks to more realistic scale-free networks.

physics.soc-ph↗