How does one maintain the look of a person on text to image generation? If I use a unique name next

How does one maintain the look of a person on text to image generation? If I use a unique name next to the type of person, shoult it theoretically keep the same identity?
Was this page helpful?