Projects & Grants




The Ability of Neural Networks to reason about language and visual data.
Project IdSGS13/PŘF-MF/2025
Main solverMgr. David Mojžíšek
Period1/2025 - 12/2025
ProviderSpecifický VŠ výzkum
Statesolved
AnotationThe project focuses on neural network research. The primary emphasis is on Large Language Models (LLMs) based on the transformer architecture. Other architectures, such as graph neural networks, are also within the scope of interest. As part of the research, we will investigate these models' reasoning capabilities, their operational efficiency, study the quality of text output, as well as identify their limitations and potential vulnerabilities.