I started by setting up the default variant in a grid with three columns and two rows. Making sure the images are centred in their frames, and no pins are active. For the "Image 1" variant, I set the zoomed-in frame to absolute position while keeping the rest in relative position. Then, added an on-click interaction from the first image to this variant.
That’s it! Just repeat the same steps for the other three variants to create the same interaction.