PhD Student (Research Assistant) at the Center of Trustworthy AI for Life Sciences - International Research Agendas Programme, University of Warsaw

Doctorat / PhD Profil chercheur : First Stage Researcher (R1) Type de contrat : Temporary Date limite de candidature : 19 Nov 2026 - 16:00 (Europe/Warsaw) Date de début : 1 Nov 2026 Programme UE : Other EU programme Référence : TRAILS-1210-12/2026 Domaine : Mathematics » OtherComputer science » OtherPhysics » Other

The person employed in the position will be responsible for: Conducting scientific research on the application of Topological Data Analysis (TDA) methods to the analysis and interpretation of how artificial neural networks function. Developing new mathematical and algorithmic methods for studying the geometry and topology of activation spaces and internal data representations within neural networks. Analyzing changes in the structure of data representations that occur during the neural network training process, and investigating their relationship to prediction quality, generalization capability, and model robustness. Developing and implementing computational tools based on TDA methods—specifically Mapper, Ball Mapper, and persistent homology—to analyze various neural network architectures, including convolutional neural networks and Transformer-based models. Designing and conducting computational experiments using real-world and synthetic datasets, including exploring the potential for identifying adversarial examples and anomalous behaviors in AI models. Preparing scientific publications, presenting research findings at conferences, and collaborating with members of an interdisciplinary research team.

The person employed in the position will be responsible for:

Conducting scientific research on the application of Topological Data Analysis (TDA) methods to the analysis and interpretation of how artificial neural networks function.

Developing new mathematical and algorithmic methods for studying the geometry and topology of activation spaces and internal data representations within neural networks.

Analyzing changes in the structure of data representations that occur during the neural network training process, and investigating their relationship to prediction quality, generalization capability, and model robustness.

Developing and implementing computational tools based on TDA methods—specifically Mapper, Ball Mapper, and persistent homology—to analyze various neural network architectures, including convolutional neural networks and Transformer-based models.

Designing and conducting computational experiments using real-world and synthetic datasets, including exploring the potential for identifying adversarial examples and anomalous behaviors in AI models.

Preparing scientific publications, presenting research findings at conferences, and collaborating with members of an interdisciplinary research team.

Where to apply E-mail [email]