59 citations
- AGH University of KrakowPL17 papers
- Czech Academy of Sciences, Nuclear Physics InstituteCZ11 papers
- Czech Academy of SciencesCZ9 papers
- Czech Technical University in PragueCZ9 papers
- University of the WitwatersrandZA9 papers
- Ghent UniversityBE7 papers
- Max Planck Institute for MathematicsDE7 papers
- Palacký University OlomoucCZ6 papers
- Jagiellonian UniversityPL5 papers
- La Trobe UniversityAU5 papers
- Shantou UniversityCN5 papers
- FZU ‒ Institute of Physics of the Academy of Sciences of the Czech RepublicCZ4 papers
110 papers
Virtual neural networks: hundreds of souls in a body
Petr Hurtik, Marek Vajgl, Zahra Alijani +1
A new concept, termed virtual neural networks, is introduced, where the count of trainable parameters is kept constant, and scalability is attained purely through computational res…
Finite-Step Bounds for Iterated Correlation Matrices
Ishrak AlhajjHassan
We establish finite-step probabilistic upper bounds on the contraction ratios for iterated Pearson correlation dynamics. Let be the sequence ge…
On limit sets and equicontinuity in the hyperspace of continua in dimension one
Domagoj Jelić, Piotr Oprocha
The paper studies the structure of -limit sets of map induced on the hyperspace of all connected compact sets, by dynamical system acting on a topolog…
First-principles discovery of stable, anisotropic, semiconducting Sb2X2O (X = S, Se) and Janus Sb2SSeO nanosheets for optoelectronics and photocatalysis
Masoud Shahrokhi, Bohayra Mortazavi
In this work, we conduct a comprehensive first-principles investigation into the design and discovery of novel antimony oxychalcogenide monolayers Sb2X2O (X = S, Se) and Janus Sb2S…
Unveiling the magnetic behavior of CN 2D material by defect creation, defect passivation, and transition metal adsorption
Taoufik Sakhraoui, František Karlický
Using the density functional tight binding method (DFTB) and the GFN1-xTB (Geometries, Frequencies, and Noncovalent interactions Tight Binding) Hamiltonian, we have investigated th…
Geometric Reasoning in the Embedding Space
Jan Hůla, David Mojžíšek, Jiří Janeček +2
In this contribution, we demonstrate that Graph Neural Networks and Transformers can learn to reason about geometric constraints. We train them to predict spatial position of point…