Cross Reference Caption Drift Visual

Create one finished image using [document captions], [cross-reference fields] and [recent inserted figures]. Depict two figure captions and their referring paragraphs connected by intended target. Check whether figure references resolve to the intended caption after insertion or reordering. Distinguish a manually typed number from an automatic reference field and do not assume updating fields fixes wrong targets. Use deliberate composition, legible visual hierarchy and no invented factual labels.

Use the attached references only for the specified visual attributes. If a required reference is missing, request it before generation. Produce one finished image with a deliberate focal point, consistent geometry and clean margins. Keep any exact wording on a separate editable label layer or leave its area blank. Do not invent evidence or present an illustration as documentary proof.