Welcome to The Salon!

The Salon is a group dedicated to the study of Statistics, Algorithms, Learning, and OptimizatioN. It is led by Professor Gautam Kamath at the Cheriton School of Computer Science at the University of Waterloo. Its members are also affiliated with the Vector Institute. Starting in Fall 2026, the group will be based in Computer Science Department of the Courant Institute of Mathematical Sciences at New York University, and will be looking for new members. Current emphases of the group include enriching our understanding of data privacy and robustness in algorithms, statistics, and machine learning.

The name of The Salon is in reference to the French practice from the 17th and 18th centuries, a central venue for exchange of some of the most important ideas of the era.

News

Recent Publications

Not All Samples Are Equal: Quantifying Instance-level Difficulty in Targeted Data Poisoning

Demystifying Foreground-Background Memorization in Diffusion Models

On the Learnability of Distribution Classes with Adaptive Adversaries

Optimal Differentially Private Sampling of Unbounded Gaussians

Avoiding Pitfalls for Privacy Accounting of Subsampled Mechanisms under Composition

Members

Le Salonneur

Avatar

Gautam Kamath

Assistant Professor of Computer Science

Statistics, Machine Learning, Data Privacy, Robustness

Graduate Researchers

Avatar

Argyris Mouzakis

PhD Computer Science Student

Machine Learning Theory, Algorithmic Statistics, Privacy, Applied Probability

Avatar

Matthew Regehr

Phd Computer Science Student

Machine Learning, Differential Privacy, Reinforcement Learning

Undergraduate Researchers

Avatar

Jack Sanderson

BA Computer Science & BA Statistics

Model Tampering, Data Poisoning, AI Security

Avatar

Valentio Iverson

BMath in Computer Science

Machine Learning Theory, Probability & Statistics, Optimization

Affiliated Researchers

Avatar

Shubhankar Mohapatra

PhD Computer Science Student

Data Privacy, Machine Learning, Federated Learning, Data Cleaning