Visualizing the Invisible presents a seminar-based course framework that integrates sensing, temporal resampling, and AI-assisted salience mapping to reveal dynamic processes that exceed human perceptual limits. Drawing parallels between transformer-based models of attention and landscape design as a relational practice, the work reframes form as the provisional outcome of intensive differences rather than static geometry. Through time-lapse imaging, frequency shifting, and abductive computational modelling, students construct salience landscapes that render environmental transformations and events perceptible. The paper contributes a reproducible workflow that expands landscape architectural representation and design agency toward temporal, process-based engagement with dynamic environments.