By harnessing physics-based machine learning, we strive to transform microscopy into a powerful tool for physics discovery. By embedding ML algorithms directly into the control systems of scanning probe and electron microscopes, the group aims to autonomously explore and manipulate materials at the atomic scale. This integration allows for rapid identification of novel physical phenomena, optimization of experimental conditions, and real-time data interpretation. Ultimately, this approach accelerates the discovery of new materials and deepens our understanding of their fundamental physical properties.”