PhD candidate/ELTE Eötvös Loránd University
Department of Artificial Intelligence/Faculty of Informatics
researching ▍
safe policy correction via targeted unlearning in reinforcement learning
reinforcement learning agents operating in dynamic environments can acquire unsafe or undesired behaviors. this work introduces safe policy correction (spc), a framework that enables targeted unlearning of specific state-action pairs without full retraining, ensuring compliance with safety constraints while preserving learned knowledge.
Dec 2025 → demoBrittle Unlearning
Interactive demo: watch published unlearning methods fail under a fine-tuning attack.
May 2026 → posta scientifically proven way to turn back time
how to organize a lan party at your ai research department, complete with a custom-built tournament manager, 3d-printed trophies, and the inevitable schedule collapse. featuring unreal tournament 2004, counter-strike 1.6, and the worms armageddon we never got to play.
Jan 2026 →