معرفی
Dr. Elisenda Grigsby is Professor of Mathematics and Computer Science at Boston College, researching geometric foundations of deep learning and topological aspects of neural networks. She directs the Experimental Math & ML Lab and teaches graduate courses on deep learning theory.
Education background:
- PhD in Mathematics, University of California at Berkeley
- AB in Mathematics, Harvard University
Her current work investigates parameter space symmetries in neural networks, functional dimension theory, and geometric representations of computational structures. She develops theoretical frameworks for transformer interpretability and studies the topological complexity of decision boundaries in deep networks.
Dr. Grigsby's publications show a transition from low-dimensional topology to ML theory, with recent work focusing on complexity measures of ReLU networks and symmetry groups in deep learning architectures. Articles increasingly incorporate tropical geometry and topological data analysis techniques.
She organizes seminars in geometry/topology and serves as editor for topology journals. Dr. Grigsby mentors students through Boston College's Math&ML graduate program and directs the Experimental Math & ML Lab.
