What does the network predict at each step?

• The noise that was added
• The photographer's name
• The file size

What does the forward process need training for?

• Nothing, it is a fixed recipe
• Choosing the image topic
• Writing captions

Generation starts from pure noise and denoises stepwise into a sample.

Circle one:   True   False

Why is stepwise denoising easier to learn than one-shot image making?

• Each step is a small easy correction
• Noise is pretty
• Computers like long loops

Samples show rough textures and odd speckles. What is the cheapest fix?

• Buy a bigger monitor
• Rename the output files
• Run more sampling steps

Generation is too slow for a demo. What do you trade away by cutting steps?

• Nothing, steps are decorative
• The need for any model
• Sample cleanliness

Two configs give equal quality but one uses half the steps. Which ships?

• The slower one, slower means better
• The half-step one, it buys the same quality faster
• Neither, quality ties prove nothing

A student trains the network to output clean images directly from pure noise. What is the conceptual error?

• Networks cannot output images
• Direct jumps skip the easy stepwise structure
• Noise must never be removed