Partha Suryanarayanan

I am a Senior Research Scientist part of Quantum Theory and Capabilities department of Algorithms and Applications division in IBM Research.

Over more than two decades, my work has ranged from building large-scale, high-performance systems in financial services to research in natural language processing, question answering, clinical text understanding and patient-record summarization, and, more recently, machine learning for molecular discovery.

My current research is in quantum machine learning. I am particularly interested in understanding when quantum representations and learning procedures provide genuinely useful inductive biases, how these effects can be separated from those of classical representation and optimization, and how quantum models should be benchmarked rigorously against strong classical alternatives.

Before moving into QML, I led research in molecular AI, including multimodal foundation models for molecular representation and generative models for small-molecule design. More broadly, I am interested in representation learning, symmetry and geometry, physics-inspired machine learning, and in developing learning methods that expose useful structure in difficult scientific problems.