Search Papers | Poster Sessions | All Posters
Poster A139 in Poster Session A - Tuesday, August 6, 2024, 4:15 – 6:15 pm, Johnson Ice Rink
The Effects of Learning on the Representational Geometry of Skilled Chess Players
Andrea Ivan Costantino1 (), Esna Mualla Gunay1, Emily Van Hove1, Laura Van Hove1, Felipe Fontana Vieira1, Merim Bilalic2, Hans Op de Beeck1; 1KU Leuven, 2Northumbria University
Chess, with its rich history as a metaphor for human intelligence, offers an excellent framework to examine expertise effects. Previous studies suggested that expert players analyse chess boards differently from novices, emphasizing piece relationships over visual traits. However, these studies did not explore representational structure and information processing changes in expertise, and in what brain areas these changes may occur. Our work bridges this gap by employing computational, behavioural, and neuroimaging methodologies to uncover representational changes in expert biological and artificial systems. By comparing chess expert and non-expert systems in humans (fMRI) and in silico (DNNs), we aim to identify chess expertise's representational changes. Our results reveal similar information processing between humans and DNNs, showing a representational and behavioural alignment between expert systems. Additionally, experts systems show a representational re-organization, resulting in more linearly separable representations of relevant high-level dimensions in late processing stages.
Keywords: learning expertise neural networks neuroimaging