Many thanks to the Julius Maximilians Universität Würzburg and Leon Bungert for hosting #MSCA researcher Albert Alcalde for two months. It lead to another #ModConFlex paper:

Quantifying Concentration Phenomena of Mean-Field Transformers in the Low-Temperature Regime

by Albert Alcalde, Leon Bungert, Konstantin Riedl, Tim Roith

➡️ https://arxiv.org/abs/2605.10931

#MachineLearning #DynamicalSystems

#MSCA #HorizonEU

ModConFlex researcher Kaja Krhač, who is an applied mathematician and theoretical physicist, spent three months at our project partner Controllab Products.

The focus of her work at Controllab was to capture the complex problem of implementing moving boundary conditions for a realistic fluid. This is a key challenge for ModConFlex signature applications such as floating wind turbines.

#MSCA
#HorizonEU

Exact Sequence Interpolation with Transformers

shows that transformer models, which power many modern AI systems, can be designed to exactly match a set of input-output examples while keeping the number of parameters under control. We prove this by showing how attention layers organize and process information efficiently, even for long sequences. Our results help explain why modern AI tools can learn, store, and handle large amounts of text so effectively.

https://arxiv.org/abs/2502.02270

#EUhorizon

#MSCA researcher Zhuo Xu published an article on the interaction between a moving particle and the surrounding fluid. He shows that the model remains well behaved.

More broadly, the paper helps improve our understanding of how fluid–structure systems can be controlled in a robust and reliable way.

IMA Journal of Mathematical Control and Information, Volume 43, Issue 2, June 2026

https://hal.science/hal-05065176v2

#ModConFlex researcher Mario Sinano co-authored

"Data-Driven Parametric Aeroelastic Modeling of the Pazy Wing", AIAA SciTech 2026.

developing physics-informed machine learning models that learn accurate, simplified representations of highly flexible aircraft wings from very small amounts of simulation data, capturing how system trajectories and stability evolve across different operating conditions.

https://arc.aiaa.org/doi/10.2514/6.2026-0187

#physics_informed #machineLearning #dynamicalsystems #aerospace

#MSCA researcher V. Ocqueteau has written the “Analysis of a Model for a Floating Platform Coupled with a Flexible Beam”. In the paper, he establishes well-posedness and derives an energy balance for the model.

The results contribute to modelling offshore applications such as offshore floating flare towers, crane vessels, meteorological masts and, most notably, wind turbines.

Currently, Vicente is staying @DLR for three months, working on simulations of the model.

https://arxiv.org/abs/2603.12786

Last week, ModConFlex researchers and professors from Besançon, Twente and Wuppertal met at the Franco-German Doctoral College at Wuppertal.

It was a great opportunity to talk and exchange with more recent and long-term members of the PHS community.

https://www.epc.ed.tum.de/rt/cdfa-phs/

#ModConFlex
#MSCA
#HorizonEU

#Modelling and #Control of Flexible #Structures Interacting with #Fluids - #Conference: 10 to 12 June 2026 at Wuppertal

Listen to and discuss talks (and present?)

➡️ geometrically nonlinear optimization of highly flexible wings
➡️ the motion of a rigid body in a perfect compressible fluid
➡️ modelling of slender locomotors at microscopic scale and applications to micro-robot control
➡️ learning provably bounded polynomial ODEs from data

More details and registration:

https://modconflex.uni-wuppertal.de/en/network-meetings/modconflex-conference/

Congratulations to the very productive #ModConFlex researchers! Over 20 papers have been written within our #MSCA project so far - several of them published.

For details, see https://modconflex.uni-wuppertal.de/en/publications

#HorizonEU