Edinburgh Napier University

  A fundamental aspect of any intelligent system is its ability to learn from experience. However, understanding how the brain develops statistical models of the world seamlessly and automatically remains an open challenge in cognitive psychology. Even though observers tend to learn simple models that are consistent with their observations over more complex but veridical models, they are not limited by environmental complexity given the right conditions. Here, I will implement a novel experimental psychophysical paradigm to measure the evolution of complex representations of visual features during learning. I will quantify how learning develops over time and reveal whether it follows a series of “eureka” moments or a continuous and smooth progress. Understanding the computational secrets of the brain will help us address issues that affect its correct function and inform our understanding of what learning is and how to facilitate it in the real world.