HKU IDS Scholar Seminar Series #23 скачать в хорошем качестве

HKU IDS Scholar Seminar Series #23

Modern machine learning models display abilities that exceed the assumptions of classical statistical learning, particularly their capacity to solve test‑time tasks far beyond those seen during training—an ability widely viewed as central to progress toward AGI. Such phenomena call for new theoretical frameworks. This talk presents a perspective based on composition: the idea that models generalize by recombining learned skills to address novel, more complex tasks. I will discuss empirical evidence and preliminary theoretical results supporting this viewpoint, aiming to motivate further investigation into this “tangram theory” of generalization.