Arthur Gretton

Arthur Gretton is a Professor at the Gatsby Computational Neuroscience Unit and director of the Centre for Computational Statistics and Machine Learning at University College London (UCL). He is also a Research Scientist at Google DeepMind, focusing on causal inference, representation learning, generative models, and nonparametric hypothesis testing.

• Research Interests: Causal modeling with hidden confounders, distributional representations via kernel methods, gradient flows for statistical divergence optimization, and robustness in machine learning.
• Recent Publications: His 2025-2024 work includes causal effect estimation with proxies, kernel-based generative model training (e.g., MMD gradient flows), and foundational work in distributional reinforcement learning.
• Scientific Recognition: Includes
  • Top 5% paper, ICLR 2023
  • Spotlight presentations at ICML 2024, NeurIPS 2023, and UAI 2023.
• Students: Current advisees include Jakub Wornbard, Zikai Shen, Zhe Wang, and others co-supervised in causal and statistical learning projects.
• Labs: Leads the Gatsby Unit and UCL Centre for Computational Statistics and Machine Learning, with software tools for kernel tests, gradient flows, and distribution embeddings.