Gautam Kamath, an Indian origin scientist has been awarded the 2026 Gödel Prize, widely regarded as the field’s most prestigious honour, for outstanding contributions to theoretical computer science.
Kamath and his colleagues, Ilias Diakonikolas, Daniel Kane, Jerry Li, Ankur Moitra and Alistair Stewart, received the award for their landmark paper, Robust Estimators in High-Dimensions without the Computational Intractability, according to a University of Waterloo release.
Awarded annually since 1993, the prize is named for Kurt Gödel in recognition of his major contributions to mathematical logic and his interest in what became known as the famous P versus NP question. The award includes a cash prize of $5,000.
Currently an Assistant Professor at the University of Waterloo’s Cheriton School of Computer Science, a Faculty Member at the Vector Institute, and a Canada CIFAR AI Chair, is set to join. the Computer Science department at the Courant Institute of Mathematical Sciences at New York University in September 2026.
READ: Indian American Neurologist Pria Anand wins literary science writing award (April 14, 2026)
Kamath is a Microsoft Research Fellow at the Simons Institute for the Theory of Computing for the Fall 2018 semester program on Foundations of Data Science and the Spring 2019 semester program on Data Privacy: Foundations and Applications.
Before that, he completed his PhD at MIT, affiliated with the Theory of Computing group in CSAIL. He graduated from Cornell University with a degree in Computer Science and Electrical and Computer Engineering in 2012.
Robust Estimators in High Dimensions Without the Computational Intractability is a landmark paper in theoretical computer science. It resolved a foundational problem that had remained elusive for decades: whether high-dimensional robust unsupervised learning can be carried out efficiently without suffering dimension-dependent degradation in accuracy.
READ: Sheel Dodani wins $100,000 Hackerman Award for protein technology research (February 26, 2026)
Before this work, understanding of the computational landscape in robust learning was limited. Even for the most basic problem of mean estimation, known statistically robust estimators were computationally intractable in high dimensions, while all known computationally efficient methods incurred error guarantees that deteriorated polynomially with the dimension.
Robust Estimators in High Dimensions without the Computational Intractability overcame this barrier. It provided the first efficient algorithms with dimension-independent error guarantees for several central high-dimensional distribution classes, and introduced a general methodology to detect and correct corruptions in high dimensions that became a blueprint for a large subsequent literature.