Postdoctoral Fellow, Multimodal Machine Learning

Develop self-supervised, multimodal models that build a continuous latent manifold of individual differences for a foundation model of human neurotypes. The fellow integrates cognitive task performance, conversational text, creative choices, psychometrics and biometric data, separates genuine cognitive variation from measurement and device effects with the project’s psychometrics postdoc, and evaluates robustness, bias and generalization.

PhD in machine learning, computer science, computational neuroscience, statistics or a related discipline; Python and modern deep-learning frameworks; HPC and GPU-cluster experience; a research background in multimodal modelling and representation learning. Self-supervised learning, NLP, cognitive modelling and privacy-aware methods are preferred. One-year contract from 1 January 2027 with possible extension; annual salary $60,000; authorization to work in Canada required.