Gids-gekoppelde quiz · Moeilijk Niveau
Memoriseren in verspreidingsmodellen Quiz
Test wat memoriseren in diffusiemodellen betekent, waarom dubbele trainingsgegevens dit veroorzaken en hoe onderzoekers bijna-kopieën extraheren en beperken.
Vraag 1 van 8
Wat is memoriseren in een diffusiemodel?
Bekijk elke vraag8 vragen
Antwoorden en uitleg verschijnen terwijl je de quiz doet.
- Wat is memoriseren in een diffusiemodel?
- The model remembering a user's previous prompts
- The model reproducing specific training images almost exactly
- The model storing images in a database for search
- The model forgetting how to draw after training
- Waarom zorgen dubbele gegevens voor een betere memorisatie?
- Duplicates are stored in higher resolution
- Duplicates are removed automatically before training
- When an image appears many times, storing it becomes the easiest way to reduce training loss on it
- Duplicates confuse the model so it outputs noise
- Wat liet het artikel uit 2023 van Carlini en collega's zien?
- Diffusion models never copy training data
- Only text models memorize data
- Memorization only happens in black and white images
- Over a hundred near-identical training images could be extracted from Stable Diffusion
- Hoe identificeerde de extractieaanval waarschijnlijk opgeslagen afbeeldingen?
- It looked for near-identical generations across many seeds for the same caption
- It read the model's weights directly as images
- It asked the model which images it had memorized
- It searched the web for similar photos
- Welke uitspraak over memoriseren is juist?
- It is the model's normal behavior for every prompt
- It is rare but can be targeted by a determined user
- It only occurs as exact pixel copies
- It has no relevance to privacy
- Why is privacy risk sharper when fine-tuning on a small dataset such as medical scans?
- Small datasets are always public
- Medical scans are too large to train on
- Each example has more influence on the model, raising the chance of regurgitation
- Fine-tuning deletes the base model
- Which mitigation bounds the influence of any single training example but currently reduces image quality?
- Differentially private training
- Increasing the number of sampling steps
- Langere aanwijzingen gebruiken
- Training for more epochs
- Why is deduplication more effective when it removes near-duplicates, not only exact copies?
- Exact copies do not affect training
- Near-duplicates are illegal to store
- Near-duplicates make images blurrier
- Slightly edited versions of the same image still act as repeated exposure