Ivan Vykopal

I'm a PhD Student in Natural Language Processing at Kempelen Institute of Intelligent Technologies (KInIT) and Brno University of Technology in Brno currently working on multilingual NLP and automated fact-checking. I previously did research at the Slovak University of Technology where I worked on applications of deep learning in medical image analysis.

My work focuses on developing AI-based methods for multilingual disinformation analysis and fact-checking. I'm particularly interested in exploring how LLMs can be used responsibly in the fact-checking process while being mindful of their potential misuse for generating disinformation. I also work on improving cross-lingual transfer learning through parameter-efficient methods like adapters and soft prompts.

I've published research on topics including multilingual claim retrieval, disinformation capabilities of large language models, and cross-lingual transfer learning. My work aims to develop practical AI tools that can help fact-checkers and journalists analyze disinformation narratives across different languages while ensuring responsible development and deployment of these technologies.