Jan Sobotka
PhD Student
Jan Sobotka is a PhD candidate in Boston University's Faculty of Computing & Data Sciences. His research aims to advance scientific understanding of deep learning, particularly language model training, so that these systems can be designed intentionally: from the curation of training data to principled architecture choices. Jan's interests span both AI and cognitive (neuro)science, especially memory and the sense of self. Before his doctoral studies, he earned a bachelor's in Informatics from the Czech Technical University in Prague and a master's in Computer Science from EPFL. Learn more about Jan.
Research focus:
- Deep learning
- Language models
- Interpretability
- Generalization