Vincent Gauthier

Modelling human activities in a system of cities Cities host most of the world population with diverse services and activities. One key challenge in urban modelling is the quantification of intra- and inter-city mobility patterns and the associated ...

Modelling human activities in a system of cities arxiv.org/abs/2602.13038

Deep Generative Model for Human Mobility Behavior Understanding and modeling human mobility is central to challenges in transport planning, sustainable urban design, and public health. Despite decades of effort, simulating individual mobility remains...

Deep Generative Model for Human Mobility Behavior arxiv.org/abs/2510.06473

Higher-order adaptive behaviors outperform pairwise strategies in mitigating contagion dynamics When exposed to a contagion phenomenon, individuals may respond to the perceived risk of infection by adopting behavioral changes, aiming to reduce their exposure or their risk of infecting others. Th...

Another new preprint on group adaptation! 📣
Great joint work w/ @martonkarsai.bsky.social and @alainbarrat.bsky.social!

Can adaptive behaviors driven by group interactions be more effective and less costly than pairwise ones? How do their adaptive mechanisms differ?

arxiv.org/abs/2602.05915

🧵 1/7

If you work on multimorbidity, network medicine, sex/gender in health, or drug repurposing, we’ve made these results (and more) available in our portal: disease-perception.bsc.es/shdc/
@alfonsovalencia.bsky.social @lmrocha.bsky.social
doi.org/10.1038/s438...

Phys. Rev. E: Influence problems on a content-spreading model and graph machine learning
http://link.aps.org/doi/10.1103/2cnl-b6wl

A Decade-long Landscape of Advanced Persistent Threats

[Blog] Between 2014 and 2023, 154 countries experienced an Advanced Persistent Threat #APT attack pulse.internetsociety.org/blog/a-decad...

Our new collective effort highlights what happens when AI acts in populations rather than as isolated tool. TL;DR: a lot.

mHC-GNN: Manifold-Constrained Hyper-Connections for Graph Neural Networks Graph Neural Networks (GNNs) suffer from over-smoothing in deep architectures and expressiveness bounded by the 1-Weisfeiler-Leman (1-WL) test. We adapt Manifold-Constrained Hyper-Connections (\mhc)~\...

arxiv.org/abs/2601.02451

Less is More: Recursive Reasoning with Tiny Networks

arxiv.org/pdf/2510.04871

Provably expressive temporal graph networks Temporal graph networks (TGNs) have gained prominence as models for embedding dynamic interactions, but little is known about their theoretical underpinnings. We establish fundamental results about th...

Provably expressive temporal graph networks
Amauri H. Souza, Diego Mesquita, Samuel Kaski, Vikas Garg

arxiv.org/abs/2209.15059

Balancing Efficiency and Expressiveness: Subgraph GNNs with Walk-Based Centrality

arxiv.org/html/2501.03...

Accelerating Bangla NLP Tasks with Automatic Mixed Precision: Resource-Efficient Training Preserving Model Efficacy Training models for Natural Language Processing (NLP) requires substantial computational resources and time, posing significant challenges, especially for NLP development in Bangla, where access to hi...

They empirically measured coordination costs, error propagation, and task structure, grounding multi-agent design from heuristics into predictive rules

more on [arXiv](arxiv.org/abs/2512.0829)

Towards a Science of Scaling Agent Systems Agents, language model-based systems that are capable of reasoning, planning, and acting are becoming the dominant paradigm for real-world AI applications. Despite this widespread adoption, the princi...

Google’s "scaling law" for building agentic systems

This paper shows when adding more AI agents helps.. and when it backfires

In a keynote at NeurIPS 2025, Melanie Mitchell argues that developmental and comparative psychologists have a lot to teach AI researchers about studying cognitive capabilities. https://spectrum.ieee.org/melanie-mitchell?share_id=9075652

screenshot of the title and authors of the Science paper that are linked in the next post

Our new article in @science.org enables social media reranking outside of platforms' walled gardens.

We add an LLM-powered reranking of highly polarizing political content into N=1256 participants' feeds. Downranking cools tensions with the opposite party—but upranking inflames them.

* Decoupling geographical constraints from human mobility *

🌍🚶 How much of our movement is about human choice and how much is constrained by geography and the spatial layout of locations?

Our paper (out in Nature Human Behavior) gives you a practical way to tell:

www.nature.com/articles/s41...

Detecting bias in algorithms used to disseminate information in social networks and mitigating it using multiobjective optimization Abstract. Social connections are conduits through which individuals communicate, information propagates, and diseases spread. Identifying individuals who a

How fair are the algorithms that decide who gets information online? Our new paper in PNAS Nexus shows that maximization algorithms, the ones used to pick “influencers” in social networks, unintentionally deepen inequality.

🧠 Sekara, Dotu, Cebrian, Moro & García-Herranz
🔗 doi.org/10.1093/pnas...

Social media feeds today are optimized for engagement, often leading to misalignment between users' intentions and technology use.

In a new paper, we introduce Bonsai, a tool to create feeds based on stated preferences, rather than predicted engagement.

arxiv.org/abs/2509.10776

Don't miss this chance!!! 🇫🇷
netmob.org/www25/
v4.event-vert.org/en/netmob2025/

QED Overview - SIGCOMM25 YouTube video by Quad9 DNS

#ICYMI the Quad9 Event Dashboard (QED) & Globe of Wonder tool were part of the Internet Visualization Exhibition (IVE) at #SIGCOMM25.

Learn more about the tool and how we use it from John's video: youtu.be/J7_Z2kROVyw?...

You can find the tool on our github github.com/Quad9DNS/Glo...

#infosec

Epidemiology models explain rumour spreading during France’s Great Fear of 1789 - Nature Epidemiological methods are used to show that the Great Fear of 1789, a series of peasant insurrections in rural revolutionary France, was driven by deliberate political action rather than spontaneous fear.

Nature research paper: Epidemiology models explain rumour spreading during France’s Great Fear of 1789

go.nature.com/4p2RjF1

Study: Social media probably can’t be fixed “The [structural] mechanism producing these problematic outcomes is really robust and hard to resolve.”…

We're probably doomed to endless toxic feedback loops unless someone hits upon a brilliant fundamental redesign that manages to change social media's core dynamics.

Just launched Sync-in — a secure, open-source platform to store, sync, and collaborate on files.

Privacy-first, self-hostable, supports real-time editing !

Designed for individuals, teams, and organizations.

👉 sync-in.com/blog/sync-in...

#opensource #selfhosted #privacy #collaboration

Scaling and Multi-scaling Laws and Meta-Graph Reconstruction in Laplacian Renormalization of Complex Networks The renormalization group (RG) method in spectral space (SS) has recently emerged as a compelling alternative to traditional RG approaches in real space (RS) and momentum space (MS). Leveraging the in...

Scaling and Multi-scaling Laws and Meta-Graph Reconstruction in Laplacian Renormalization of Complex Networks arxiv.org/abs/2507.08893

The recent events show Egypt's very poor ISP market competition and dependence on a dominant local upstream provider is affecting its overall Internet resilience. Read our analysis pulse.internetsociety.org/blog/egypt-i...

Slide titled 'network effects in scientific labor', arguing that your own productivity is a network effect, because it includes the productivity of your coauthors too

Slide showing a worked example of a collaboration network, and using a probabilistic model to marginalize the network effects of collaboration to estimate individual productivity (without network effects)

Slides showing results of applying the network effect model showing that collaborating with a highly productive or highly prominent coauthor in your early career has a lasting effect on your own productivity and prominence, suggesting that social capital (network effects) are a kind of semi-transferrable form of wealth

Slides describing a network analysis of homophily in collaboration networks, showing that diversity in collaborators is a learned tendency, and mid-career researchers who trained in highly diverse research groups go on to lead relatively more diverse groups themselves

Slides from my keynote lecture "What drives the productivity of scientific labor?" at the 2025 Oxford Summer School on Economic #Networks. Part 2 of 2: how does who you collaborate with shape your own productivity, and how are networks like social capital? aaronclauset.github.io/slides/Claus...

Initialisation and Network Effects in Decentralised Federated Learning Fully decentralised federated learning enables collaborative training of individual machine learning models on a distributed network of communicating devices while keeping the training data localised ...

Initialisation and Network Effects in Decentralised Federated Learning arxiv.org/abs/2403.15855

Universal expansion of human mobility across urban scales Human mobility is a fundamental process underpinning socioeconomic life and urban structure. Classic theories, such as egocentric activity spaces and central place theory, provide crucial insights int...

Universal expansion of human mobility across urban scales arxiv.org/abs/2406.06889

Fareed’s Take: Trump is gutting what made American science great YouTube video by CNN

Farreed's take: youtu.be/0Qk4FBS73Ts?...

Using 🐁watching "The Matrix", scientists have reconstructed the largest and most detailed wiring diagram of a mammalian brain to date, by mapping cells in a cubic millimetre of a mouse’s brain tissue.

82k neurons, 500 million synapses

#Neuroscience 🧪🧠

www.nature.com/articles/s41...