Deep Learning Advanced Quiz 1

1. Which architecture is the basis for most modern large language models (LLMs)?

2. What is the main purpose of attention mechanisms in deep learning?

3. Which regularization technique randomly sets some activations to zero while training?

4. What is a key benefit of using residual connections in deep networks?

5. Which optimizer combines momentum with adaptive learning rates?

6. What is the main idea behind GANs (Generative Adversarial Networks)?

7. Which loss function is commonly used when training standard GANs?

8. In GANs, what is “mode collapse”?

9. Which method is commonly used to interpret complex deep learning models?

10. What is the main purpose of layer normalization?