Local Causal Attribution of Chain-of-Thought Reasoning
- 2026
- ICML 2026
Workshop paper
Yannis Belkhiter is a Pre-Doctoral AI Researcher at IBM Research Dublin and Trinity College Dublin, advised by Dr. Seshu Tirupathi and Prof. John D. Kelleher.
His research interest lies on Natural Language Processing. His thesis focuses on Large Reasoning Models, specifically investigating the internal structure of reasoning traces, with the goal of making their generation more efficient, accurate, and secure. Yannis is also interested in the Security of Machine Learning, and Mechanistic Interpretability.
Prior to start his PhD program, Yannis earned an MSc in Computing from Imperial College London, and completed an MEng in Computer Science at École des Mines d'Alès.