AIML Rounds

QUIZ

Generative models

8 questions on Generative models, from 6 lessons. Pick an answer in your head, then open it.

On the roadmap for: ML Engineering.

  1. Question 1 of 8

    What does an embedding learn, for example for foods like biryani and pulao?

    1. A. A vector where similar items end up close together
    2. B. The number of letters in each item's name
    3. C. A random ID number for each item
    4. D. A single category label for each item, stored as a one-hot vector
    View answer

    ✓ A. A vector where similar items end up close together

    Embeddings place similar items near each other, so similarity can be measured with distances or cosine.

    Lesson: Embeddings and Representation Learning

  2. Question 2 of 8

    What makes a variational autoencoder able to generate new data, unlike a plain autoencoder?

    1. A. It has more layers than a plain autoencoder
    2. B. It skips the decoder when generating new data
    3. C. A smooth latent space tied to a prior, for sampling
    4. D. It uses class labels to guide the decoder
    View answer

    ✓ C. A smooth latent space tied to a prior, for sampling

    The KL term keeps the latent space well organised, so random samples decode into plausible outputs.

    Lesson: Autoencoders and VAEs

  3. Question 3 of 8

    Why are GANs hard to train?

    1. A. They have no loss function at all for the optimiser to work on
    2. B. Two competing networks can oscillate or collapse
    3. C. They can't be trained on GPUs
    4. D. They need labelled data for every image
    View answer

    ✓ B. Two competing networks can oscillate or collapse

    The generator and discriminator must stay balanced; tricks like better losses and normalisation help.

    Lesson: GANs

  4. Question 4 of 8

    How does a diffusion model generate an image?

    1. A. From pure noise, it removes a little noise each step
    2. B. It draws the image pixel by pixel, row after row
    3. C. It copies and blends training images
    4. D. It uses a GAN discriminator to pick images
    View answer

    ✓ A. From pure noise, it removes a little noise each step

    Training adds noise to real images and learns to undo it; generation runs that denoising many times.

    Lesson: Diffusion Models

  5. Question 5 of 8

    How did CLIP learn to match images and text without hand-labelled classes?

    1. A. By translating captions into other languages
    2. B. Matching image–caption pairs scored above mismatches
    3. C. From manual labels for every image
    4. D. By predicting the pixels of each image from the caption's words
    View answer

    ✓ B. Matching image–caption pairs scored above mismatches

    Contrastive learning pulls matching pairs together and pushes others apart, enabling zero-shot classification.

    Lesson: Contrastive Learning (SimCLR, CLIP)

  6. Question 6 of 8

    In contrastive learning, what are the 'negatives'?

    1. A. Images with negative pixel values
    2. B. Examples with wrong labels in the dataset
    3. C. Other examples in the batch that don't match
    4. D. Training runs that failed to converge
    View answer

    ✓ C. Other examples in the batch that don't match

    More and harder negatives usually make the learned representation more useful.

    Lesson: Contrastive Learning (SimCLR, CLIP)

  7. Question 7 of 8

    Which family usually gives the most stable training and best image quality today, at the cost of slower generation?

    1. A. Diffusion models
    2. B. GANs
    3. C. Plain autoencoders
    4. D. Linear regression
    View answer

    ✓ A. Diffusion models

    Diffusion trains stably and covers the data well; it needs many denoising steps, which faster samplers reduce.

    Lesson: Diffusion Models

  8. Question 8 of 8

    What is the intersection over union (IoU) of two 10×10 boxes that overlap in a 5×10 area?

    1. A. About 0.25
    2. B. About 0.33
    3. C. Exactly 1.0
    4. D. About 0.5
    View answer

    ✓ B. About 0.33

    Intersection 50, union 100 + 100 − 50 = 150, so IoU = 50 / 150 ≈ 0.33.

    Lesson: Deep Learning Practice