Erfan Hatefi

Hi, I’m Erfan Hatefi, though my full name is Sayed Mohammad Vakilzadeh Hatefi (but don’t worry, you can stick with Erfan). I’m currently a Research Assistant specializing in eXplainable AI at the Fraunhofer Heinrich Hertz Institute, under the supervision of Sebastian Lapuschkin and Wojciech Samek. I work closely with Max Dreyer and Reduan Achtibat, who are invaluable collaborators and mentors.

My research currently focuses particularly on uncovering explainable subgraphs within vision and text-based models (Mechanistic Interpretability). Additionally, I have an interest and expertise in model compression (especially through pruning) using explainability techniques, and in exploring the relationship between robustness and XAI.

I’m pursuing a Master’s degree in Computer Science at TU Berlin. Prior to this, I completed my Bachelor’s degree at the University of Kashan in Iran, where I graduated as the ranked-first student under the supervision of Saeed Asaeedi and Mahsa Soheil Shamaee.