Yes, what you're observing is normal in computational chemistry, especially during Intrinsic Reaction Coordinate animations. When you are studying a ring-closing reaction using the QST2 method in Gaussian, the motion of atoms particularly hydrogens can often appear messy. This typically happens because the IRC traces the minimum energy path from the transition state to both the reactants and products, and in doing so, even subtle energetic preferences or steric effects get reflected in the atomic trajectories. Hydrogen atoms on CH₂ groups may appear to interact with the carbon involved in ring formation, but in most cases, these interactions are not chemically significant. Instead, they are often due to transient geometric adjustments, might be van der Waals contacts, or simply visualization artifacts that emerge as the molecule passes through compressed transition geometries. Unless these hydrogen interactions result in consistent close contacts or lead to changes in energy profiles or reaction barriers, they are usually not of chemical importance.