Jakub Kwiatkowski

Jakub Kwiatkowski

Auto-generated profile, based on publicly available information

Jakub Kwiatkowski

Institute of Computing Science, Poznan University of Technology, Poznan, Poland

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 March 2024 in Security and Defence Quarterly
Verified
Journal Article
Open access
PDF file
MTMateusz Grzegorz Twardawa
Mateusz Grzegorz Twardawa
MSMarek Smolik
Marek Smolik
FRFranciszek Rakowski
Franciszek Rakowski
JKJakub Kwiatkowski
Jakub Kwiatkowski
NMNorbert Meyer
Norbert Meyer
SCADvanceXP is an industrial network intrusion detection system that scans and monitors data exchange between engineering stations, field divides, controllers, supervisory control and data acquisition (SCADA), and other elements of the operational technology network in detail. SCADvanceXP has the...

23 May 2025 in IEEE Transactions on Neural Networks and Learning Systems
Verified
Journal Article
Closed access
JKJakub Kwiatkowski
Jakub Kwiatkowski
KKKrzysztof Krawiec
Krzysztof Krawiec
This study presents and investigates abstract compositional transformers (ACTs), a class of deep learning (DL) architectures based on the transformer blueprint, designed to handle abstract reasoning tasks that require completing spatial visual patterns. We combine ACTs with choice-making modules and apply them to Raven progressive matrices (RPMs), logical puzzles that require selecting the correct image from the available answers. We devise a number of ACT variants, train them in several modes and with additional augmentations, subject them to ablations, demonstrate their data scalability, and analyze their behavior and latent representations that emerged in the process. Using self-supervision allows us to successfully train ACTs on relatively small training sets, mitigate several biases identified in RPMs in past studies, and achieve SotA results on the two most popular RPM benchmarks.

3 August 2024 in Security and Defence Quarterly
Verified
Journal Article
Open access
MTMateusz Grzegorz Twardawa
Mateusz Grzegorz Twardawa
MSMarek Smolik
Marek Smolik
FRFranciszek Rakowski
Franciszek Rakowski
JKJakub Kwiatkowski
Jakub Kwiatkowski
NMNorbert Meyer
Norbert Meyer
Industrial infrastructure has suffered an unprecedented number of attacks in Central and Eastern Europe (CEE). This situation can be attributed to many geopolitical factors, including hybrid military conflicts and criminal activity. Industrial networks belonging to the countries that were once under Soviet influence suffer from an elevated risk of cyberattacks. The goal of this work is to propose an easy way to estimate the vulnerabilities of industrial networks to cyber threats on a national level. Since analysis of the industrial vulnerability landscape is difficult, this study proposes an assessment based on the popularity of vulnerable technologies—VPc. This metric is composed of search volume data on keywords related to industrial network technologies and reported security vulnerabilities associated with these words. Data on 116 keywords was analysed and a country-specific VPc index was calculated for twenty states in CEE. The analysis of the popularity of industrial technologies and vendors in CEE reveals interesting information about the industrial security and vulnerability landscape. The results show that some countries (e.g. Estonia) have more resilient industrial infrastructure than others (e.g. Belarus). The results presented in this study are not in conflict with other data and estimation attempts, including the National Cyber Security Index (NCSI). As new vulnerabilities are noted every day, the industrial security landscape changes rapidly. Therefore, a new easy-to-use metric (VPc) can be successfully used for general estimations. This work shows that the VPc score agrees with other estimates and analyses, but as with any other general estimation tool, it must be used with caution.