Ahmed Elhag

I am a PhD student at the Department of Computer Science at the University of Oxford, advised by Michael Bronstein. My research focuses on the intersection of geometric deep learning, graph ML, and generative models. I’m particularly interested in how we can combine these approaches to develop robust ML methods that can accelerate progress in drug discovery and molecular design. One of my current research areas is learning symmetry and equivariance in unconstrained models to address the limitations of strictly equivariant architectures (see Relaxed Equivariance via Multitask Learning and Learning Inter-Atomic Potentials without Explicit Equivariance).

Before my doctoral studies, I graduated from the African Masters of Machine Intelligence program at AIMS Senegal. I also interned at Apple MLR team where I worked on developing generative models for 3D and graph-structured data.

Contact: ahmed.elhag[at]cs.ox.ac.uk