Hi There 👋 , I'm Mehrdad Farahani
A fourth-year PhD student at Chalmers and the University of Gothenburg, working on memory arbitration (including internal behaviors) in LLMs. My work connects mechanistic interpretability and controllability, aiming to understand how models internalize, retrieve, and adapt, and how these processes can be guided in practice. I also have formal training in applied project management for leading collaborative research and engineering efforts.