Statistical learning is the brain's ability to notice patterns and predict future events based on these patterns. Using brain imaging technology (fMRI), we studied how different parts of the brain work together to support statistical learning by examining their connections. Participants completed a task designed to test statistical learning. We then measured their learning performance and examined how it related to changes in brain connectivity. We found that better learning was linked to stronger activity in certain brain areas: the lateral occipital cortex, angular gyrus, precuneus, anterior cingulate cortex, and superior frontal gyrus. Our analysis showed that the activity in the superior frontal network was most strongly related to learning performance. Furthermore, connections between the superior frontal gyrus and several other brain networks (related to attention, language, and salience) were weaker during the learning task. This weakening of connections may help the superior frontal network focus on identifying new patterns by reducing interference from other processes. We suggest that these weaker brain connections allow for better processing of new information, enhancing the ability to learn from the environment.