This model was trained on photos of interpretive signs that had been obliterated with tagging on a nearby nature walk.

Why? Nobody has any use for such a thing!

Because I wanted to see what would happen. All sorts of strange effects can occur in SDXL, while in Juggernaut it tends to create backgrounds. But that's just with relatively short prompts and weight near 1.