How far can the political views of language models be steered?
A paper on how far system-prompt instructions can steer the answers of large language models was presented at AIES 2026 in Malmö.
News
Visiting researcher Dr. Wojciech Cygan from the University of Wrocław
Dr. Wojciech Cygan (University of Wrocław) visited FER and worked with Stjepan Šebek on fluctuation laws of random walks with drift.
VALOR at the European Frontier AI-Safety Day in Berlin
The principal investigator took part in a meeting of European frontier AI safety researchers.
Better bounds for the "inradius" of Brownian motion
Stjepan Šebek and co-authors improved the bounds on the inradius of the convex hull of planar Brownian motion.
VALOR at the 4th Korean–Croatian Summer Probability Camp in Dubrovnik
Stjepan Šebek took part in the Korean–Croatian summer probability camp in Dubrovnik.
Stable long-horizon prediction of partial differential equation solutions
At Equadiff 2026 in Prague, Domagoj Vlah presented latent-tensor vision transformers for fast and stable prediction of parametric PDE solutions.
VALOR at the "Cognitive Sovereignty" workshop in Paris
The principal investigator took part in an invited HES workshop at Institut Polytechnique de Paris.
The convex hull of a Brownian bridge with a random endpoint
A new paper gives closed formulas for the expected perimeter and area of the convex hull of a family of processes between Brownian bridge and Brownian motion.
Fast simulations on domains of changing shape
A paper in Computers & Mathematics with Applications presents reduced order models that use deep neural networks to learn the domain description on their own.
Perimeter and area of the Brownian convex hull in the unit disk
Stjepan Šebek and co-authors computed the exact expected perimeter and area of the convex hull of Brownian motion stopped upon exiting the unit disk.
VALOR project launched
A four-year project on interpretability and value alignment of large language models started at FER on 1 October 2025.