Deep Learning Advanced Quiz 3

1. Which method is often used to stabilize GAN training?

2. What is the main function of the decoder in an encoder–decoder model?

3. Which regularization technique penalizes large weights by adding their squares to the loss?

4. What is a “bottleneck layer” in deep architectures?

5. Which of the following is a challenge in training deep RNNs?

6. What is a main advantage of the GELU activation function in Transformers?

7. Which loss function is typically used in neural style transfer?

8. What is the main idea behind curriculum learning?

9. Which technique helps generalization by combining predictions from several models?

10. What is “knowledge distillation” in deep learning?