Themes
The CommBank Centre for Foundational AI 糖心volg advances core research for foundational AI. The Centre is focused on three distinct themes:
Theme 1: Bias, trust, and attribution in deploying reliable AI models
This theme focuses on developing AI models that are not only accurate but also fair, transparent, and accountable. It will explore biases within model architectures and learning algorithms, methods for model interpretability and explainability, and techniques for ensuring trust and reliability in real-world deployment.
Theme 2: Responsible reasoning and forecasting in deep learning
This theme is dedicated to understanding when and how AI models exhibit genuine reasoning capabilities. It aims to develop systems that can forecast future outcomes, infer structured conclusions from incomplete or uncertain data, and uncover the underlying principles that govern their reasoning processes.
Theme 3: Scalable and resource-efficient models for large-scale AI
This theme will develop foundations for building high-performing AI models that are computationally and memory efficient, enabling deployment at scale. It includes research into model compression, low-rank and quantised architectures, efficient training algorithms, and energy-aware model design.