Ludlow, Kiera2025-05-072025-05-072025https://hdl.handle.net/1828/22168This study investigates category learning, a process where individuals progress from basic novel recognition of stimuli or categories to expert-level understanding, enabling quicker and more accurate categorization. To explore this, we are utilizing behavioural data, like reaction time and accuracy, and psychological embeddings from PsiZ, a machine learning tool. Furthermore, we are using RUBubbles, a novel artificial stimuli designed to avoid any participant perceptual biases that may arise from prior familiarity. Our study involved 10 participants in a pre- and post-test design. Participants initially rated the similarity of novel images, completed a same-different species identification task, and were trained to distinguish four RUBubble families. A post-test conducted one week later evaluated how well participants retained and retrieved category knowledge. Accuracy, and reaction times were examined to assess changes in neural and behavioral performance. Additionally, PsiZ was used to visualize how participants' psychological representations of the categories evolved through training. Results displayed that training resulted in improved behavioral performance and significantly different category structures in the psychological embeddings. These findings enhance our understanding of how perceptual training and feedback shapes category representations, offering valuable insights into the neural mechanisms underlying learning.category learningpsychological embeddingspsizneural correlatesrububblesPoster