I am a researcher at SFB TRR318 “Constructing Explainability”, a Collaborative Research Center at Paderborn University. I am working in Project A05 alongside its principal investigators, Prof. Dr. Katharina J. Rohlfing and Prof. Dr.-Ing. Britta Wrede.

In my PhD, I worked on extending contrastive explanations, a fundamental method in explainable AI, from one-shot decision settings to interactive human–AI explanatory tasks. This work combined insights from psycholinguistics, computational linguistics, cognitive science, explainable AI, and human–robot interaction.

I am also interested in understanding multimodal processes in human–AI interaction and how they can contribute to generating more understandable explanations. To this end, I am currently working to bring established methods from psycholinguistics and cognitive science, such as visual-world paradigms, into HRI and XAI research to investigate language-vision interaction in real time during human–AI explanation tasks.

Updates

🎉 Paper accepted for publication in the Journal of Psycholinguistic Research

Our paper A Method for Studying Online Language–Vision Interplay in Human–Robot Interaction has been accepted for publication in the Journal of Psycholinguistic Research. Accepted manuscript

🎉 PhD successfully defended 🎓

I successfully defended my doctoral dissertation, From Negation to Contrast and Context: A Framework for Contrastive Guidance of Actions in Interaction. Link

🎉 Paper accepted for presentation at IEEE ICDL 2026

Our paper A Divergence Model of Scaffolding in Dialogue via Negation has been accepted for presentation at the IEEE International Conference on Development and Learning (ICDL 2026). The study explores how people build shared understanding during conversation. In conversations, two people may begin with slightly different understandings of a task. The paper shows that negation—for example, clearly stating what something is not—can make these differences visible and guide the conversation back on track. Preprint

🎉 Paper accepted for publication in Acta Psychologica

A recent paper analyzing eye-tracking data from a preferential-looking paradigm has been accepted for publication in Acta Psychologica. The study shows how individual differences in infants’ preferences for salient object shapes influence word learning. Link

The open-access book on Social Explainable AI (sXAI) is now out. Read

I contributed a chapter on Explanation Goals. Link

TeaP 2026, Tübingen

I recently presented my work at TeaP 2026 in Tübingen. Read the abstract

IEEE ICDL 2025, Prague

At IEEE ICDL 2025 in Prague, I gave a talk on language-vision interaction in human-robot interaction. Watch the talk on YouTube

WinGaze

WinGaze is a tool for exploring multimodal interaction in HRI. View on GitHub