Quiz | Croma Campus

Which algorithm is used for generating text data in a sequential manner?

LSTM
K-Means
PCA
SVM

Mark for review

Which technique is used to improve the training stability of GANs by slowing down the learning rate of the discriminator?

Curriculum Learning
Wasserstein Distance
Adversarial Training
Spectral Normalization

Mark for review

Which Generative AI model uses a probabilistic approach to represent data?

Variational Autoencoders (VAEs)
Generative Adversarial Networks (GANs)
Autoencoders
Transformer models

Mark for review

Which layer is typically used in the generator of a GAN to upsample data?

Convolutional Layer
Pooling Layer
Transposed Convolutional Layer
Dense Layer

Mark for review

What does the term "mode collapse" refer to in the context of GANs?

The discriminator becomes too strong
The generator fails to learn any meaningful patterns
The generator produces a limited variety of outputs
The training process becomes unstable

Mark for review

Which technique is used for conditioning GANs to generate specific outputs?

Dropout
One-hot encoding
Batch normalization
Label smoothing

Mark for review

Which of the following is NOT a type of generative model?

GAN
VAE
RNN
PixelCNN

Mark for review

Which of the following is a disadvantage of Generative Adversarial Networks (GANs)?

They are difficult to train.
They cannot generate diverse outputs.
They are prone to mode collapse.
They require large amounts of labeled data.

Mark for review

Which of the following techniques is used for generating realistic images with controllable attributes?

StyleGAN
DCGAN
CycleGAN
Pix2Pix

Mark for review

Which technique is used to improve the diversity of generated samples in a Generative AI model?

Increasing the learning rate
Decreasing the batch size
Adding noise to the inputs
Using feature matching

Mark for review

Which technique is used to combine features from multiple layers in a Generative AI model?

Skip connections
Data augmentation
Gradient descent
Dropout

Mark for review

In VAE, what is the role of the decoder?

To generate data
To compress input data into a latent space
To evaluate the generated data
To reconstruct the original input data from the latent space

Mark for review

Which of the following is NOT a characteristic of generative models?

Supervised learning
Unsupervised learning
Semi-supervised learning
Reinforcement learning

Mark for review

Which loss function is commonly used in training Variational Autoencoders (VAEs)?

Mean Squared Error (MSE)
Cross Entropy Loss
Reconstruction Loss
Discriminator Loss

Mark for review

Which of the following is a popular dataset used for training Generative AI models?

MNIST
ImageNet
CIFAR-10
All of the above

Mark for review

Which of the following techniques is commonly used in generative adversarial networks (GANs)?

Support Vector Machines (SVM)
Reinforcement Learning
Backpropagation
Variational Autoencoders (VAEs)

Mark for review

Which of the following is a limitation of VAEs?

Difficulty in generating high-quality images
Unstable training process
Limited expressiveness in latent space
All of the above

Mark for review

Which of the following is NOT a type of GAN architecture?

DCGAN
WGAN
VGGAN
SAGAN

Mark for review

What is the primary objective of generative AI?

To generate data
To generate new content similar to a given dataset
To classify data
To optimize existing models

Mark for review

Which Generative AI model is typically used for generating high-resolution images?

Autoencoders
Variational Autoencoders (VAEs)
Generative Adversarial Networks (GANs)
Transformer models

Mark for review

Which Generative AI model is based on a "teacher-student" framework?

Generative Adversarial Networks (GANs)
Variational Autoencoders (VAEs)
Generative Pre-trained Transformers (GPT)
Autoencoders

Mark for review

Which technique can be used to interpolate between two images in latent space?

Style transfer
Linear interpolation
Gradient descent
Random sampling

Mark for review

Which of the following is a benefit of using GANs for data augmentation?

Reduced computational cost
Improved model generalization
Faster training convergence
Increased interpretability of the model

Mark for review

Which of the following statements is true regarding GANs?

GANs require labeled data for training
GANs are only used for image generation
GANs consist of a single neural network
GANs can suffer from mode collapse

Mark for review

Which method is used to interpolate between two points in the latent space of a Generative AI model?

Principal Component Analysis (PCA)
Linear interpolation
Gradient descent
K-means clustering

Mark for review

In a GAN, what is the role of the generator?

To generate fake data
To evaluate the generated data
To classify data
To optimize the discriminator

Mark for review

Which of the following techniques is used for semi-supervised learning with Generative AI models?

Label smoothing
Adversarial training
Dropout
Self-training

Mark for review

In VAE, what is the purpose of the encoder?

To generate data
To compress input data into a latent space
To evaluate the generated data
To optimize the decoder

Mark for review

Which loss function is typically used in training a GAN?

Mean Squared Error (MSE)
Cross-Entropy Loss
Hinge Loss
KL Divergence

Mark for review

Which of the following is a popular framework for training Generative Adversarial Networks (GANs)?

TensorFlow
Scikit-learn
PyTorch
Keras

Mark for review

Online Exam Quiz

Online

Did you like this Question

Get a 1:1 Mentorship call with our Career Advisor