Date and time: 25 July 2024, 11am to 1pm
Talk location: Redmond Barry Building, Level 12, 1201
Light refreshments location: Redmond Barry Building, Level 12 pantry
Talk details:
Title
Reading minds: Using generative modelling to reveal the brain’s hidden code
Abstract
Our experience of the world is shaped by dynamic neural computations that emerge over a range of timescales (evolutionary, learning, adaptation) and are concealed from conscious awareness. These computations are central to dominant theories of brain function, such as Bayesian inference, yet their nature and implementation remain elusive. To understand how and why the brain alters sensory information to support adaptive behaviour and cognition, we must reveal these hidden codes. In this presentation, I will provide two examples of how we can begin to address this question by combining generative modelling with neural decoding. In the first example, I will show how we can characterise the shape of prior beliefs about the environment from neural recordings and how this can be implemented within the brain in a manner consistent with Bayesian inference. In the second example, I will show how we can use the same method to measure experimentally manipulated changes to these beliefs, revealing novel insights into old phenomena, i.e., sensory adaptation.
Speaker bio:
Dr Reuben Rideaux is a Senior Lecturer and NRHMC Emerging Leader at the University of Sydney's School of Psychology. Prior to this he was a DECRA Fellow at the Queensland Brain Institute, a Leverhulme Early Career Fellow at the University of Cambridge and a PhD student at the Australian National University. He combines computational modelling, brain imaging, and psychophysics to study perception and cognition. He has a particular interest in developing new methods for understanding brain function and dysfunction, such as bio-inspired artificial intelligence systems, high resolution functional MR spectroscopy, and neural decoding.
