Naturally. That's how LLMs work. During training you measure the loss, the difference between the model output and the ground-truth and try to minimize it. We prize models for their ability to learn. Here we can see that the large model does a great job at learning to draw bob, while the small model performs poorly.

For the model to have memorized the entire sequence of characters precisely, this must appear hundreds of times in the training data?

Conveniently removed the artist's signature though.