Susan Wei
Associate Professor
Department of Econometrics and Business Statistics, Monash University
Biography
Susan Wei is an Associate Professor in the Department of Econometrics and Business Statistics at Monash University. She previously held an Australian Research Council Discovery Early Career Researcher Award (DECRA, 2020–2023) and was a visiting faculty researcher at Google DeepMind in Sydney (2023).
Her research is in probabilistic machine learning, with a current focus on Bayesian predictive inference. She works on the martingale posterior and related predictive constructions of uncertainty, including limit theory for learned predictive rules such as transformers and prior-data fitted networks. She applies these tools to the probabilistic interpretability of meta-learned sequence models, treating a trained sequence model as an implicit predictive rule and recovering the prior and posterior it encodes. She has also worked on Bayesian deep learning, variational inference, and singular learning theory.
She is a 2026 recipient of a UK AI Security Institute Alignment Project grant for the project Probabilistic Interpretability of Transformers, and has received unrestricted research gifts from Google (2022, 2026), the most recent of which also supports this line of work.
Interests
- Bayesian predictive inference
- Martingale posteriors
- Interpretability of sequence models
- Tabular foundation models
- Singular learning theory
Education
PhD in Statistics, 2014
University of North Carolina, Chapel Hill
BA in Mathematics, 2009
University of California, Berkeley