About

Three-line bio

Five-line bio

Sune Lehmann is a Professor of Networks and Complexity Science at the Technical University of Denmark (DTU) and Professor of Social Data Science at the University of Copenhagen. A physicist by training, his research uses massive datasets and approaches from complex systems, machine learning, and network science to understand patterns in human behavior. His work spans human mobility, complex contagion, sleep, collective attention, and the prediction of life trajectories, with publications in Nature, Nature Physics, PNAS, and Nature Computational Science, among others. He led the Copenhagen Networks Study, one of the largest social sensing experiments in Europe, and his life2vec project — which uses transformer models to predict life events from population-scale registry data — received worldwide media coverage in outlets from Science to The Washington Post. He is a member of the Royal Danish Academy of Sciences and Letters, recipient of Denmark’s EliteForsk Senior Prize, and Chief Scientist in the Danish National Center for AI in Society.

One-third page bio

Sune Lehmann is a Professor of Networks and Complexity Science at DTU Compute, Technical University of Denmark, and Professor of Social Data Science at the Center for Social Data Science (SODAS), University of Copenhagen. A physicist by training — with a B.Sc. and M.Sc. from the Niels Bohr Institute and a Ph.D. from DTU — his research draws on complex systems, machine learning, and network science to develop a quantitative understanding of human behavior from large-scale data.

Before returning to Denmark, Lehmann held postdoctoral positions at Northeastern University’s Center for Complex Network Research, the Dana-Farber Cancer Institute at Harvard, and Harvard’s Institute for Quantitative Social Science. His research has produced high-impact contributions across a wide range of topics: community detection in networks (Nature, 2010), human mobility patterns (Nature, 2020; Nature Human Behaviour, 2018), the accelerating pace of collective attention (Nature Communications, 2019), and most recently, using transformer models to predict life events from population-scale data (Nature Computational Science, 2024).

Lehmann led the Copenhagen Networks Study, one of Europe’s largest social sensing experiments, tracking the social interactions, mobility, and communication patterns of over a thousand university students. He served on the Danish government’s COVID-19 modeling task force and the advisory board for Denmark’s Smittestop contact tracing app. He is a member of the Royal Danish Academy of Sciences and Letters, a recipient of the EliteForsk Senior Prize, and Chief Scientist in the Danish National Center for AI in Society (CAISA). He has supervised 12 PhD students, 14 postdocs, and more than 75 Master’s projects.

Half-page bio

Sune Lehmann is a Professor of Networks and Complexity Science at DTU Compute, Technical University of Denmark, and Professor of Social Data Science at the Center for Social Data Science (SODAS), University of Copenhagen. A physicist by training — with degrees from the Niels Bohr Institute and a Ph.D. from DTU on the structure of complex networks — his research sits at the intersection of physics, computer science, and the social sciences, using massive datasets and methods from complex systems to study how people behave, move, connect, and influence one another.

After completing his doctorate, Lehmann spent three years in Boston as a postdoctoral fellow, splitting time between Albert-László Barabási’s Center for Complex Network Research at Northeastern University, the Center for Cancer Systems Biology at Harvard’s Dana-Farber Cancer Institute, and Harvard’s Institute for Quantitative Social Science. He returned to DTU in 2010 and has maintained yearly research visits to Northeastern ever since.

Lehmann’s work has produced contributions that have shaped several areas of network science and computational social science. His 2010 Nature paper on link communities introduced a method for detecting overlapping community structure in networks that has become a standard tool in the field. His group’s work on human mobility — showing that people maintain a stable number of frequently visited locations throughout their lives, reallocating rather than accumulating — was published in Nature (2020) and Nature Human Behaviour (2018). Research on the accelerating dynamics of collective attention (Nature Communications, 2019) documented how the window of public focus on cultural topics has been shrinking across decades and media, from books to Twitter. More recently, the life2vec project, published in Nature Computational Science (2024), applied transformer models to sequences of life events drawn from Danish population registries, demonstrating that the language-model paradigm can capture deep structure in human life trajectories. The work generated worldwide press coverage, from Science and The Washington Post to The Economist and AFP syndication across dozens of countries.

From 2012 to 2019, Lehmann led the Copenhagen Networks Study (previously SensibleDTU), one of the most comprehensive social sensing experiments in Europe, equipping over a thousand students with instrumented smartphones to capture face-to-face interactions, communication, mobility, and sleep — producing a dataset that has underpinned dozens of publications across the group. He is co-editor of the Springer volume Complex Spreading Phenomena in Social Systems (2018) and has delivered keynotes at IC2S2, NetSci, Complex Networks, and Digital Tech Summit, among others.

Lehmann served on the Danish government’s COVID-19 modeling task force and the advisory board for the national Smittestop contact tracing app, and was part of the expert group advising the Danish government on regulation of tech giants. He is a member of the Royal Danish Academy of Sciences and Letters, a recipient of the EliteForsk Senior Prize (2022), co-recipient of the Columbus Prize (2021), and was part of the team recognized as Best Danish Research Environment (2023). He is currently Chief Scientist in the Danish National Center for AI in Society (CAISA) and serves as an editor for PNAS. At DTU, he teaches courses on social graphs and data visualization to hundreds of students each year, and has been nominated for best teacher four times and best PhD advisor once. He has supervised 12 PhD students, 14 postdocs, and more than 75 Master’s projects.

One-page bio

Sune Lehmann is a Professor of Networks and Complexity Science at DTU Compute (Technical University of Denmark) and Professor of Social Data Science at the Center for Social Data Science (SODAS), University of Copenhagen. He is also Chief Scientist in the Danish National Center for AI in Society (CAISA) and a member of the Royal Danish Academy of Sciences and Letters. A physicist by training, his research draws on complex systems, machine learning, and network science to build a quantitative understanding of human behavior from large-scale data.

Lehmann grew up in Denmark and studied physics at the Niels Bohr Institute, University of Copenhagen, completing his B.Sc. in 2000 and M.Sc. in 2003. His master’s thesis, supervised by Benny Lautrup, explored complex networks and scientific excellence — a topic that foreshadowed two decades of research. He earned his Ph.D. from DTU in 2007, with a dissertation on the structure of complex networks supervised by Lars Kai Hansen.

After his doctorate, Lehmann spent three years in Boston. He held simultaneous postdoctoral positions at Albert-László Barabási’s Center for Complex Network Research at Northeastern University and the Center for Cancer Systems Biology at Harvard’s Dana-Farber Cancer Institute, followed by a fellowship at Harvard’s Institute for Quantitative Social Science. These years immersed him in a remarkable community of network scientists and cemented long-running collaborations — with Barabási, David Lazer, Sandy Pentland, Yong-Yeol Ahn, Alan Mislove, and many others — that have shaped his career. He has returned to Northeastern for summer research visits nearly every year since.

Lehmann joined the DTU faculty in 2010, was promoted to Associate Professor in 2012 and Full Professor in 2019. He also holds an Adjunct Professorship in Social Network Science at the Department of Sociology, University of Copenhagen.

His research spans a striking range of topics unified by a common thread: revealing hidden patterns in large-scale behavioral data. In community detection, his 2010 Nature paper (with Ahn and Bagrow) introduced the concept of link communities, a method for uncovering overlapping structure in networks that has become widely used across disciplines. In human mobility, his group has shown that people maintain a fixed “capacity” of regularly visited places, swapping locations in and out rather than accumulating them (Nature, 2020; Nature Human Behaviour, 2018). Work on collective attention demonstrated that the public’s attention span on cultural phenomena — from bestselling books to trending hashtags — has been systematically shrinking over decades (Nature Communications, 2019). His research on complex contagion, using carefully controlled experiments with Twitter bots, provided some of the first experimental evidence that information can spread through social reinforcement rather than simple exposure (PLOS One, 2017; Best Paper Award, IC2S2 2018).

From 2012 to 2019, Lehmann led the Copenhagen Networks Study, one of the most ambitious social sensing experiments in Europe. The project equipped over a thousand university students with instrumented smartphones, capturing face-to-face interactions via Bluetooth, communication patterns, GPS mobility traces, and even sleep behavior — generating a high-resolution dataset that has underpinned dozens of publications on topics from friendship formation to epidemic modeling.

Most recently, the life2vec project (Nature Computational Science, 2024) applied transformer architectures — the technology behind large language models — to sequences of life events drawn from Danish national registries, demonstrating that these models can learn meaningful representations of human lives and predict future outcomes. The work generated a wave of international media coverage, from Science and The Washington Post to Reuters and AFP, and prompted important public conversations about the ethical boundaries of prediction.

Lehmann’s work has been covered extensively in international media throughout his career. The TwitterMood project was featured on CBS Evening News, in The New York Times, Scientific American, BBC, TIME, and The Wall Street Journal. The Economist, Bloomberg, TechCrunch, and Mother Jones have covered his research on mobility and social bots. He has contributed op-eds and public commentary on topics ranging from tech regulation to contact tracing to the societal impact of AI.

He served on the Danish government’s COVID-19 modeling task force and advised on Denmark’s national contact tracing app. He was part of an expert group advising the Danish government on regulating tech giants, and serves on the executive board of the Network Science Society. He is an editor for PNAS and an academic editor for PLOS One and Applied Network Sciences.

As a teacher, Lehmann has developed and taught courses on social networks, data visualization, and computational social science at DTU since 2010, reaching hundreds of students annually. He was an early adopter of flipped-classroom and project-based methods, and has been nominated for DTU’s best teacher award four times. He has supervised 12 PhD students, 14 postdocs, and more than 75 Master’s projects. His former students and postdocs hold positions at Google, UNICEF, Northeastern University, and as faculty across Danish and European universities.

Lehmann has been recognized with the EliteForsk Senior Prize (2022) — one of Denmark’s most prestigious research awards — the Columbus Prize (2021), and was part of the team named Best Danish Research Environment (2023). He was elected to the Royal Danish Academy of Sciences and Letters in 2024 and is a member of the Danish Academy of Technical Sciences.

He is co-editor of Complex Spreading Phenomena in Social Systems (Springer, 2018), and has received over DKK 75 million in research funding as PI or Co-PI, including a Villum Synergy Grant, a Carlsberg Semper Ardens grant, an ERC Advanced Grant, and a Sapere Aude Starting Grant. He has delivered keynotes at IC2S2, Complex Networks, Digital Tech Summit, and TechBBQ, and given invited talks at institutions including MIT, Stanford, Harvard, Cambridge, ETH Zurich, the Santa Fe Institute, and the Royal Danish Academy.

Three-page bio

I’m a physicist who ended up studying people.

The short version is that I’m a Professor of Networks and Complexity Science at DTU (the Technical University of Denmark) and Professor of Social Data Science at the University of Copenhagen. But the longer version is more interesting — it’s about how a kid who loved physics gradually discovered that the most fascinating complex systems aren’t made of particles or proteins. They’re made of people.

How I got here

I studied physics at the Niels Bohr Institute in Copenhagen, the place where Bohr and Heisenberg had their famous conversations about quantum mechanics. My bachelor’s thesis was on quantum entanglement. My master’s thesis, supervised by Benny Lautrup, was on complex networks and scientific excellence — basically, I was using network science to understand how scientific careers work. That was 2003, and network science was still a young field. I was hooked.

For my Ph.D. at DTU (completed in 2007), I went deeper into the structure of complex networks. I worked on community detection — the problem of identifying groups and clusters in large networks — and developed methods that would eventually lead to one of my most-cited papers. My advisors were Lars Kai Hansen, Andrew Jackson, and Mogens Høgh Jensen, a combination of machine learning, physics, and complex systems that pretty much defined the interdisciplinary approach I’ve taken ever since.

After my doctorate, I moved to Boston for what turned out to be three formative years. I worked simultaneously at Northeastern University — in Albert-László Barabási’s Center for Complex Network Research — and at the Dana-Farber Cancer Institute at Harvard, applying network methods to biological systems. Later I moved to Harvard’s Institute for Quantitative Social Science, where I worked with David Lazer on computational social science before the field even had that name.

Boston was also where I met many of my closest collaborators and friends in science: Yong-Yeol Ahn, Jim Bagrow, Alan Mislove, Sandy Pentland. Some of my best work has come from those friendships. I’ve gone back to Northeastern for summer research visits nearly every year since — more than a decade of annual returns.

I joined the DTU faculty in 2010, became an Associate Professor in 2012, and was promoted to Full Professor in 2019. Since 2020, I’ve also been Professor of Social Data Science at the University of Copenhagen.

What I work on

My research is about finding hidden patterns in large-scale data about human behavior. I started in network science — the mathematics of connections — and gradually moved toward using networks as a lens for understanding social systems. Along the way, I’ve worked on a perhaps surprising range of topics, but they’re all connected by a common thread: using data and computational methods to understand things about people that we couldn’t see before.

Networks and communities. My 2010 Nature paper with Yong-Yeol Ahn and Jim Bagrow introduced the concept of link communities — the idea that in real networks, it’s the relationships, not the individuals, that belong to communities. Each person sits at the intersection of multiple overlapping groups (family, work, friends), and you can only see that structure clearly if you focus on the links. The algorithm we developed has been used across fields from biology to social science.

Human mobility. With Laura Alessandretti, Piotr Sapieżyński, Andrea Baronchelli, and others, I’ve spent years studying how people move through space. One of our key findings, published in Nature in 2020, is that people maintain a roughly constant number of regularly visited locations throughout their lives — around 25. When we adopt a new place, we drop an old one. It’s like a social Dunbar’s number, but for places. We also showed that mobility patterns are organized in a nested hierarchy of scales, from the neighborhood coffee shop to cross-country trips.

Collective attention. With Philipp Lorenz-Spreen, Bjarke Mønsted, and Philipp Hövel, I studied how the dynamics of public attention have changed over time. Our 2019 Nature Communications paper showed that across multiple cultural domains — from Twitter hashtags to Google Books to movie ticket sales to Reddit — the window of collective attention has been shrinking. Topics rise faster and fall faster. The nine-day wonder that Chaucer wrote about is now a six-hour wonder.

Complex contagion. With Bjarke Mønsted, Piotr Sapieżyński, and Emilio Ferrara, I ran one of the first controlled experiments on information spreading using Twitter bots. We showed that information can spread through social reinforcement — you need to hear something from multiple sources before you’ll share it — providing experimental evidence for a phenomenon that had been theorized but never cleanly demonstrated.

The Copenhagen Networks Study. Starting in 2012, I led one of the most ambitious social sensing experiments in Europe. We gave over a thousand DTU students instrumented smartphones that recorded their face-to-face interactions (via Bluetooth), communication patterns, GPS traces, app usage, and more. The resulting dataset — with resolution down to five-minute intervals — has been the foundation for dozens of papers across my group, on topics from friendship formation to sleep patterns to epidemic modeling. Flashing the custom firmware onto a thousand phones in 2013 remains one of my fondest (and most exhausting) memories.

Life trajectories and life2vec. My most recent major project, led by Germans Savcisens and published in Nature Computational Science in 2024, applied transformer architectures to sequences of life events from Danish national registries — education, employment, health, income — treating an entire human life as something like a sentence in a language model. The model learned meaningful representations of life trajectories and could predict future events. The work generated enormous press coverage and raised important questions about what it means to “predict a life.” We were careful to frame it as a research contribution about the structure of human experience, not a tool for deployment. There is a fine line between understanding and surveillance, and I care a great deal about staying on the right side of it.

Teaching and mentoring

I’ve taught courses on social networks, data visualization, and computational social science at DTU since 2010, typically reaching 200–350 students a year across two courses. I’m a strong believer in project-based, hands-on learning — my students collect data from real APIs, build real visualizations, and work with real networks. I was an early adopter of flipped-classroom methods at DTU, and I’ve been nominated for best teacher four times (haven’t won yet — but I keep trying).

I’ve supervised 12 PhD students, 14 postdocs, and more than 75 Master’s projects. Watching my students succeed is genuinely one of the best parts of my job. Former group members hold positions at Google, UNICEF, Northeastern University, and as faculty at universities across Denmark and Europe. Several have started companies. One of the things I’m most proud of is that many of my former students stay in touch and continue to collaborate long after they leave the group.

Public life

I believe strongly that scientists have a responsibility to engage with the public. My work has been covered in The New York Times, The Washington Post, BBC, Scientific American, The Economist, TIME, CBS Evening News, Le Monde, and many other outlets. I’ve been involved in public debates about tech regulation, contact tracing, and the societal implications of AI. I served on the Danish government’s COVID-19 modeling task force, advised on the national contact tracing app (where I pushed hard for a privacy-preserving decentralized architecture), and was part of an expert group advising the Danish government on how to regulate tech giants.

Since 2025, I serve as Chief Scientist in the Danish National Center for AI in Society (CAISA). I was elected to the Royal Danish Academy of Sciences and Letters in 2024 and received the EliteForsk Senior Prize in 2022 — one of Denmark’s most prestigious research awards.

I also care about the scientific community itself. I was on the executive board of the Network Science Society for over a decade, served as General Chair of NetSci 2013 in Copenhagen (for which we received an event award from the mayor), and I’m an editor for PNAS and PLOS One. I’ve reviewed grants for the NSF, the European Commission, and the Volkswagen Foundation, among others.

The bigger picture

What drives me is the conviction that computational approaches can help us understand social systems in genuinely new ways — but only if we’re careful about how we do it. The data we work with represents real people living real lives. The patterns we find can be used to help or to harm. I try to do work that matters, that’s methodologically rigorous, and that respects the people behind the data points.

My Erdős number is three, in case you were wondering.

Five-page bio

I’m a physicist who ended up studying people.

That sentence has been my elevator pitch for years, and it still captures the essence of what I do. I’m a Professor of Networks and Complexity Science at DTU (the Technical University of Denmark) and Professor of Social Data Science at the University of Copenhagen. I’m also Chief Scientist in the Danish National Center for AI in Society. But titles only tell you what someone is called, not what they actually do all day — so here’s the longer version.

Education and early career

I grew up in Denmark and studied physics at the Niels Bohr Institute in Copenhagen — the same building where Bohr, Heisenberg, and Pauli once argued about the foundations of quantum mechanics. It’s a place that takes ideas seriously, and I loved it. My bachelor’s thesis (2000) was on quantum entanglement, which sounds like it belongs in a different career, but there’s a thread connecting it to what I do now: even then, I was drawn to systems where the relationships between things matter more than the things themselves.

For my master’s thesis (2003), supervised by Benny Lautrup, I made the leap from quantum systems to complex networks, studying the structure of citation networks and scientific excellence. Network science was still a young field — Barabási and Albert’s paper on scale-free networks was only four years old — and I was captivated by the idea that you could use mathematics and data to reveal hidden structure in social systems.

My Ph.D. at DTU (completed 2007) went deeper into this territory. My dissertation, “The Structure of Complex Networks,” supervised by Lars Kai Hansen with co-advisors Andrew Jackson and Mogens Høgh Jensen, tackled the problem of community detection: how do you identify meaningful groups in large, messy networks? The combination of advisors — machine learning, nuclear physics, and biophysics — gave me an interdisciplinary foundation that has defined my approach ever since.

After my doctorate, I moved to Boston for three years that fundamentally shaped my career. I held overlapping postdoctoral positions at Northeastern University, in Albert-László Barabási’s Center for Complex Network Research, and at Harvard’s Dana-Farber Cancer Institute, where I applied network methods to biological data. Later I moved to a fellowship at Harvard’s Institute for Quantitative Social Science, working with David Lazer on what would come to be called computational social science.

Boston was where I found my scientific community. I met Yong-Yeol Ahn, Jim Bagrow, and Alan Mislove at Northeastern; Sandy Pentland at MIT; and many others who have remained close collaborators and friends ever since. Some of my best work has grown directly out of those relationships. I’ve returned to Northeastern for summer research visits almost every year for more than a decade — a tradition that has become one of the fixed points in my calendar.

I joined the DTU faculty in 2010, was promoted to Associate Professor in 2012 and Full Professor in 2019. Since 2020, I’ve held a joint professorship at the University of Copenhagen’s Center for Social Data Science (SODAS), and since 2017, an adjunct professorship in the Department of Sociology there as well.

Research themes

My research is about using data and computational tools to understand human behavior at scale. I’ve worked on a wide range of topics — networks, mobility, sleep, attention, contagion, academic performance, life trajectories — but they’re all connected by a shared conviction: that there are deep, quantifiable patterns in how people live, and that finding those patterns can genuinely change how we understand ourselves.

Networks and community detection

This is where it all started. The central question in network science is how to identify meaningful structure in complex systems — the groups, clusters, and communities that organize a network. My early work developed methods for modularity optimization and biclique communities. But the breakthrough came in 2010, when Yong-Yeol Ahn, Jim Bagrow, and I published a paper in Nature introducing the concept of link communities.

The key insight is simple but powerful: in real networks, communities overlap pervasively. Every person belongs to multiple groups — family, colleagues, college friends, neighbors. Traditional methods force each node into a single cluster, which misses this fundamental reality. Our approach was to assign links, not nodes, to communities. Your connection to your sister belongs to the “family” cluster; your connection to your colleague belongs to “work.” But you belong to both.

The paper has been cited thousands of times, and the method has been applied across biology, social science, and information science. I think the reason it worked is that it captures something true about how the social world is organized.

Human mobility

With Laura Alessandretti, Piotr Sapieżyński, Andrea Baronchelli, Marta González, and others, I’ve spent a decade studying how people move through physical space. We’ve shown that human mobility is far more structured than it appears.

Our 2018 paper in Nature Human Behaviour demonstrated that people maintain a conserved quantity in their mobility: a roughly fixed number of regularly visited locations at any given time. When you adopt a new restaurant, gym, or coffee shop, you tend to drop an old one. It’s like a Dunbar’s number for places.

The 2020 Nature paper (with Alessandretti and Ulf Aslak) extended this, showing that human mobility is organized as a nested hierarchy of spatial scales — from your daily neighborhood routines up through city-level patterns to long-distance travel. The structure is remarkably consistent across individuals and datasets. This work has implications for urban planning, epidemiology, and our understanding of how people relate to the spaces they inhabit.

Earlier work in the group also pioneered methods for tracking mobility from WiFi signals rather than GPS — showing that you can reconstruct someone’s movements with surprising accuracy just from the access points their phone scans for. The privacy implications of that finding remain important.

Collective attention and information dynamics

With Philipp Lorenz-Spreen, Bjarke Mønsted, and Philipp Hövel, I studied how the dynamics of collective attention have changed over time. Our 2019 Nature Communications paper documented something that many people feel intuitively but that hadn’t been rigorously measured: across multiple cultural domains — Twitter hashtags, Google Books, movie ticket sales, Reddit posts, and more — the window of collective attention has been shrinking. Topics rise faster, peak higher, and disappear sooner. The nine-day wonder that Chaucer wrote about has become a six-hour wonder.

This work ended up as part of the discussion around Johann Hari’s book Stolen Focus — an experience that taught me a lot about the difference between careful academic discourse and the rough-and-tumble of public debate on Twitter. I wrote about that episode on my blog, and I still think the paper’s findings are important: not because they tell us anything about individual attention spans, but because they reveal something about how our collective focus is structured by the information systems we’ve built.

Complex contagion

How does information spread through social networks? The simple model is that it works like a disease: one exposure is enough to “infect” you. But social influence is often more complex than that. You might need to hear about something from multiple independent sources before you take it seriously.

With Bjarke Mønsted, Piotr Sapieżyński, and Emilio Ferrara, I ran one of the first controlled experiments on this question, using Twitter bots to expose real users to URLs in a carefully designed setup. We found clear evidence of complex contagion: people were significantly more likely to share content when they’d been exposed to it by multiple sources. The paper (PLOS One, 2017) won the Best Paper Award at IC2S2 2018 and remains one of the cleanest demonstrations of social reinforcement in information spreading.

I also co-edited the book Complex Spreading Phenomena in Social Systems (Springer, 2018) with Yong-Yeol Ahn, bringing together contributions from across the field.

The Copenhagen Networks Study

From 2012 to 2019, I led one of the most ambitious social sensing experiments ever conducted in Europe. The Copenhagen Networks Study (previously called SensibleDTU) equipped over a thousand DTU students with instrumented smartphones that recorded Bluetooth proximity (who was physically near whom), communication logs, WiFi-based location, accelerometer data, and more — all at five-minute resolution, for years at a time.

I still remember the logistics of it: flashing custom firmware onto a thousand phones in 2013, debugging sensor issues, navigating privacy regulations, and managing the sheer complexity of keeping the whole system running. The dataset it produced has been extraordinary — a rare, high-resolution window into the social lives of a large population, enabling research that simply isn’t possible with surveys or social media data alone.

The study has underpinned dozens of publications from my group and many collaborators. We used it to study friendship formation, the relationship between physical proximity and social ties, class attendance and academic performance, sleep patterns, daily rhythms, epidemic spreading, and more. The data from the project has been published as a resource for the broader research community.

Sleep and daily rhythms

An unexpected direction that grew out of the Copenhagen Networks Study was sleep research. With Sigga Svala Jónasdóttir and James Bagrow, I’ve studied how sleep patterns vary across gender, age, and context using data from wearable devices, published in Nature Human Behaviour (2022) and Sleep (2021). We found, among other things, that people naturally compensate for sleep debt while traveling — a hopeful finding for anyone who has ever slept beautifully in a hotel room. We also showed systematic gender differences in sleep patterns and variability across the adult lifespan using a global-scale wearables dataset.

Academic performance

With Andreas Bjerre-Nielsen, Valentin Kassarnig, and David Dreyer Lassen, I pursued a trilogy of papers on predicting academic performance using behavioral data from the Copenhagen Networks Study. The punchline — and it genuinely surprised us — was that simple administrative data (past grades) outperformed the rich behavioral data we’d painstakingly collected. The surveillance-style “big data” didn’t add predictive power beyond what you could learn from a transcript.

The implication, which I find deeply important, is that for many practical applications, we should think carefully about whether we actually need to collect all that behavioral data. As we put it in the PNAS paper (2021): the right data often beats big data. It’s an argument for privacy, for parsimony, and for thinking carefully about what information is truly relevant to the question you’re trying to answer.

Life trajectories and life2vec

My most recent major project, led by Germans Savcisens and published in Nature Computational Science in 2024, asked a provocative question: can you treat the sequence of events in a person’s life — education, employment, health, income, relationships — as something like a language, and use the same techniques that power large language models to learn its structure?

The answer is yes. We trained a transformer model on life-event sequences from Danish national registries, and it learned rich, meaningful representations of human lives — capturing relationships between education, health, social position, and many other dimensions. The model could predict future life events, including health outcomes and mortality risk.

The project generated an enormous wave of media coverage — Science, The Washington Post, The Economist, AFP (syndicated to dozens of countries), and many more. Much of the viral coverage sensationalized the work as a “death predictor,” which is not what it is. We built a clarification page (life2vec.dk) to set the record straight. The work is about understanding the deep structure of human experience, not about deploying prediction tools on individuals. The difference between understanding and surveillance matters, and I care a great deal about getting it right.

Scientometrics and the science of science

A persistent thread in my work has been studying science itself. My 2006 Nature correspondence (with Andrew Jackson and Benny Lautrup) on measures of scientific quality was an early contribution to the debate about bibliometrics and research evaluation. Our 2018 PNAS paper on “the chaperone effect” (with Vedran Sekara, Pierre Deville, Sebastian Ahnert, Barabási, and Roberta Sinatra) showed that first-time authors publishing in prestigious journals typically do so with an experienced co-author who has published there before — a kind of mentorship structure embedded in the publication system.

Teaching and mentoring

I’ve taught courses on social networks, data visualization, and computational social science at DTU since 2010, reaching 200–350 students per year. My two main courses — “Social Graphs and Interactions” and “Social Data Analysis and Visualization” — are project-based and hands-on: students collect real data through APIs, analyze networks and text, build visualizations, and present findings. I was an early adopter of flipped-classroom methods at DTU, and I’ve been nominated for best teacher four times.

I’ve supervised 12 PhD students and 14 postdocs, plus more than 75 Master’s projects. I’m also nominated for best PhD advisor. Mentoring is one of the parts of my job I take most seriously. Watching a student go from tentative first steps to independent, confident researcher is genuinely one of the most rewarding experiences in academic life. My former students and postdocs hold positions at Google, UNICEF, Northeastern University, the IT University of Copenhagen, DTU, and universities across Europe. Several have started companies (including Peergrade). Many of them remain close collaborators and friends.

Press and public engagement

I believe scientists have a responsibility to communicate their work to the public. Over the years, my research has been covered in The New York Times, The Washington Post, Scientific American, The Economist, BBC, CBS Evening News, TIME, Vanity Fair, Le Monde, NPR, and many other outlets. The TwitterMood work in 2010 generated a particularly large wave, including a segment on CBS Evening News with Katie Couric and multiple pieces in the Times. The life2vec project in 2023–2024 produced another major cycle of international coverage.

Beyond press coverage, I’ve been active in public debate. I’ve written op-eds in Danish media on tech regulation and AI, served on expert groups advising the government, and spoken at public events from Science and Cocktails in Christiania (still my favorite talk I’ve ever given — the audience actually laughed at my jokes) to the Royal Danish Academy. My blog at sunelehmann.com has been a space for longer-form thinking — about filter bubbles, about the difference between big data and the right data, about the ethics of prediction.

Awards and recognition

I’ve been fortunate to receive recognition for my work. The EliteForsk Senior Prize (2022) is one of Denmark’s most prestigious research awards. The Columbus Prize (2021) recognized contributions to science communication. In 2023, my research environment was named Best Danish Research Environment. I was elected to the Royal Danish Academy of Sciences and Letters in 2024 and have been a member of the Danish Academy of Technical Sciences since 2021.

Service and leadership

Since 2025, I serve as Chief Scientist in CAISA, the Danish National Center for AI in Society. I’m Associate Director of SODAS and Collaboratory Lead at the Pioneer Center for AI. I served on the executive board of the Network Science Society for over a decade, was General Chair of NetSci 2013 in Copenhagen, and I’m an editor for PNAS, PLOS One, Applied Network Sciences, and Frontiers of Physics. I review grants for the NSF, the European Commission, ANR, and the Volkswagen Foundation, and review papers for Nature, Science, PNAS, and many other journals.

I served on the Danish government’s COVID-19 modeling task force, the advisory board for the Smittestop contact tracing app, and an expert group on the regulation of tech giants. These roles have reinforced my view that scientists can and should contribute to public policy — carefully, humbly, and with an awareness of the limits of our models.

Research funding

My work has been supported by major grants including a Villum Foundation Synergy Grant (DKK 19.7 million), a Carlsberg Foundation Semper Ardens Grant (DKK 25 million, co-PI), an ERC Advanced Grant (co-PI), a Sapere Aude Starting Grant from the Danish Council for Independent Research, a Villum Young Investigator Grant, and funding from the University of Copenhagen’s Programme of Excellence for Interdisciplinary Research. Total funding as PI or co-PI exceeds DKK 75 million.

The bigger picture

What drives me is the conviction that computational approaches can help us understand social systems in genuinely new ways — but only if we do it with care. The data I work with represents real people living real lives. The patterns we find can be used to inform policy, improve public health, and deepen our understanding of ourselves — or they can be used for surveillance, manipulation, and control. The difference is in the intention, the methods, and the institutional safeguards we build around the work.

Richard Feynman once responded to the Whitman poem about the astronomer by arguing that understanding nature scientifically doesn’t diminish its beauty — it adds new layers of wonder. I feel the same way about human behavior. The more I learn about the hidden patterns in how we move, connect, and influence each other, the more remarkable the whole thing seems.

My Erdős number is three, by the way. I’m also technically on IMDb — I appeared in an episode of the TV show Be Sharp about life2vec (“Computers Can Predict When You’re Going to Die… Here’s How,” 2024). So my Erdős-Bacon number exists in principle, though I haven’t managed to calculate the Bacon half yet.