Peer collaboration has long been a cornerstone of scientific advancement, and nowhere is this tenet truer than in network science, where the entire premise of the field holds that a complex system is far more than the sum of its parts. It's a principle the AccelNet-MultiNet Fellowship Program takes seriously, and literally. Today, as a final 2025–2026 cohort of Fellows prepares to depart for partner research hubs across the Atlantic, the program stands as one of the most focused efforts to accelerate the science of multilayer networks, embedding graduate students, postdoctoral researchers, and early career faculty at institutions with complementary expertise in the field.
"This exchange program offers an opportunity to learn from—and collaborate with—new colleagues, and to initiate new connections between US and European researchers and institutions," explains the Network Science Institute Executive Director, Kate Coronges, who spearheaded this new round of fellowships. "AccelNet-MultiNet Fellows join a community of scholars who have created their own ecosystem of activities, research and leadership in this field."
This idea took root in 2020 based on an ambitious vision: what if the scattered pockets of expertise in multilayer network science, distributed across universities and research institutes in the U.S. and Europe, could be woven together into a multilevel knowledge network of its own? That vision became the AccelNet-MultiNet Program, funded by the NSF and anchored at Northeastern University's Network Science Institute. Today, with 44 international exchanges completed, it has more than delivered on that early mission.
The premise
Fellows—graduate students, postdoctoral scholars, and early career faculty in the area of network science who focus on multilevel networks— spend one to four months embedded at a partner institution across the Atlantic. Long enough to do real work. Long enough to form real relationships. Europe's strengths in theoretical modeling, network geometry, and temporality meet America's expertise in structural analysis and network dynamics, accelerating new ideas that would have taken much longer for either side to achieve alone.
This model places AccelNet-MultiNet in a distinguished tradition of international exchange programs like the Erasmus Programme, launched by the European Union in 1987, and the Fulbright Program, running since 1946. What distinguishes AccelNet-MultiNet from these broader initiatives is its deliberate focus. Rather than facilitating exchange across all disciplines or all geographies, it concentrates resources on a single, rapidly evolving scientific frontier —the theory of multilevel networks — and on a carefully selected constellation of institutions that together cover the full intellectual landscape of that field.
Partners & participants
Scholars from U.S., at the Networks Science Institute in Boston and Portland, ME, and Indiana University (Indianapolis, IN) have the opportunity to travel to one of the European Accelnet partner hubs at the University de Barcelona (Spain), ISI Foundation (Turin, Italy), Central European University (Vienna, Austria), Center for Theoretical Physics (Marseilles, France), or to NetSI’s London campus, and vice-versa. Fifteen Fellows are participating in this year’s cohort; 13 of them are graduate students.
Jesseba Fernando, a PhD student at NetSI Boston, kept running into the same question throughout her work in systems neuroscience and AI interpretability: how do networks reorganize when they learn something new? “One of the central mysteries in neuroscience is how the brain learns to flexibly recombine things it already knows” —something researchers call compositional generalization, a mystery Jesseba is aiming to solve drawing on network theory, dynamical systems, and neuroscience. “No single lab has the full picture, but working with Giovanni Petri [NetSI London], Andrea Brovelli [Institute de Neurosciences de la Timone], and Alain Barrat [Center for Theoretical Physics in Marseilles] gives me access to exactly that combination. The AccelNet fellowship felt like a rare opportunity to pursue compositionality in a structured way.”
As someone who helped build YouTube in its early days, NetSI PhD student Hong Qu, on the other hand, is drawing on that experience to explore through the Fellowship one of the unintended consequences of getting news via social media. “At NetSI, I study how people access news and information and how that shapes their political views. At Trinity College [a new AccelNet partner] I’m excited to work with Prof. Taha Yasseri’s group to build a multilayer dynamic network to simulate how people's news sources shifted from mainstream journalism to social media influencers. Being in Dublin—home to many major tech companies—is an added bonus and a chance to connect with industry.”
For José Andrés Guzmán Morán, PhD student at NetSI London, the AccelNet Fellowship comes at a particularly critical moment in his PhD. “I have been developing two main research lines: inference of contact patterns from epidemic data, and modeling contagion processes with group interactions. Until now, these two directions have progressed largely in parallel. Although I had already planned to design a project that would bring them together, the opportunity to pursue this integration through the AccelNet Fellowship significantly changes the scope and feasibility of the project.” Jose will be working closely with Guillaume St-Onge, a Research Assistant Professor at NetSI Portland who is tackling closely related problems. “Although we are part of the same institute, physical distance clearly affects how easily collaboration happens, and the AccelNet Fellowship has emerged as a concrete way to build bridges between different hubs.”
This intercontinental and multi-institutional collaborative effort ultimately harnesses the power of people to advance a collective scientific agenda. Yet perhaps equally lasting is the human lattice it leaves behind, researchers who return home not just with new data or tools, but with co-authors, friendships, and a firsthand understanding of how science is done differently elsewhere.
Funding for the AccelNet-Multinet Program was provided by the National Science Foundation (Award #1927425)
