Unit 5 - Practice Quiz

INT344 50 Questions
0 Correct 0 Wrong 50 Left
0/50

1 What is the primary characteristic of a Sequence Model compared to a standard Feedforward Neural Network?

A. It takes the order of inputs into account and can handle variable-length inputs
B. It uses only Convolutional layers
C. It assumes all inputs are independent of each other
D. It processes inputs of fixed length only

2 Which of the following data types is best suited for a Sequence Model?

A. Iris flower categorization
B. Tabular housing price data
C. Static image classification
D. Sentiment analysis of movie reviews

3 In a Recurrent Neural Network (RNN), what is the function of the 'hidden state'?

A. To store the final output class
B. To act as a memory that captures information about previous time steps
C. To reset the network weights after every epoch
D. To visualize the attention weights

4 What is the phenomenon called when the gradients become extremely small during the backpropagation through time in an RNN, preventing weights from updating?

A. Vanishing Gradient
B. Gradient Clipping
C. Overfitting
D. Exploding Gradient

5 Which algorithm is typically used to train Recurrent Neural Networks?

A. Standard Backpropagation
B. K-Means Clustering
C. Random Forest
D. Backpropagation Through Time (BPTT)

6 Which activation function is most commonly used for the hidden state in a simple RNN to help regulate values?

A. Softmax
B. Linear
C. ReLU
D. Tanh

7 What is the primary architectural solution designed to solve the Vanishing Gradient problem in standard RNNs?

A. Long Short-Term Memory (LSTM)
B. Convolutional Neural Network (CNN)
C. Autoencoder
D. Perceptron

8 In an LSTM unit, which gate is responsible for deciding what information to discard from the cell state?

A. Forget Gate
B. Update Gate
C. Output Gate
D. Input Gate

9 What represents the 'long-term memory' component in an LSTM architecture?

A. Output Gate
B. Input Gate
C. Cell State (C_t)
D. Hidden State (h_t)

10 In an LSTM, what is the range of values output by the sigmoid activation function used in gates?

A. -1 to 1
B. -infinity to +infinity
C. 0 to 100
D. 0 to 1

11 Which task involves assigning a grammatical category (like Noun, Verb, Adjective) to every word in a sentence?

A. Part-of-Speech (POS) Tagging
B. Sentiment Analysis
C. Machine Translation
D. Named Entity Recognition

12 Named Entity Recognition (NER) is primarily concerned with identifying:

A. Sentiment of the text
B. Real-world objects like people, organizations, and locations
C. The translation of the text
D. Grammatical errors

13 What type of Sequence problem is POS Tagging?

A. Many-to-Many (Synced)
B. Many-to-One
C. One-to-Many
D. One-to-One

14 In the context of NER, what does the 'BIO' or 'IOB' tagging scheme stand for?

A. Beginning-Inside-Outside
B. Binary-Input-Output
C. Basic-Input-Operation
D. Backward-Inward-Onward

15 What is the core architecture used in Neural Machine Translation (NMT) before the introduction of Attention?

A. Encoder-Only
B. Encoder-Decoder (Seq2Seq)
C. Random Forest
D. Decoder-Only

16 In a traditional Seq2Seq model, what is the role of the Encoder?

A. To generate the output sequence
B. To calculate the loss function
C. To visualize the data
D. To compress the input sequence into a fixed-length context vector

17 What is the 'Context Vector' in a traditional RNN-based Encoder-Decoder model?

A. The weights of the output layer
B. The average of all input vectors
C. The last hidden state of the encoder
D. The first hidden state of the encoder

18 Which of the following is a major bottleneck of the traditional Seq2Seq model?

A. It can only translate into English
B. Performance degrades significantly for long sentences due to the fixed-length context vector
C. It cannot handle numeric data
D. It requires too much RAM

19 What is 'Teacher Forcing' in the context of training sequence models?

A. Manually setting the weights of the network
B. Forcing the model to stop training early
C. Using the model's predicted output as input for the next step during training
D. Using the actual ground truth output from the previous time step as input for the current step during training

20 Which search strategy explores multiple possible output sequences simultaneously to find the most likely translation?

A. Beam Search
B. Linear Search
C. Greedy Search
D. Binary Search

21 The Attention Mechanism was primarily introduced to solve which problem?

A. The inability of RNNs to process images
B. The information bottleneck of the fixed-length context vector in NMT
C. Slow training of Linear Regression
D. Overfitting in CNNs

22 How does the Attention Mechanism calculate the context vector for each time step in the decoder?

A. By computing a weighted sum of all encoder hidden states
B. By averaging all input words
C. By taking the last state of the encoder only
D. By randomly selecting an encoder state

23 In Attention, what do the 'alignment scores' (or attention weights) represent?

A. The number of hidden layers
B. The error rate of the model
C. How relevant a specific input word is to the word currently being generated
D. The magnitude of the gradient

24 What mathematical function is typically applied to alignment scores to convert them into probabilities that sum to 1?

A. Softmax
B. Sigmoid
C. Tanh
D. ReLU

25 In the context of RNNs, what is 'Backpropagation Through Time' (BPTT)?

A. Training the network in reverse order
B. Unfolding the RNN across time steps and applying backpropagation
C. Using future data to predict past data
D. A method to predict future stock prices

26 Which of the following is NOT a gate in a standard LSTM?

A. Output Gate
B. Attention Gate
C. Forget Gate
D. Input Gate

27 What is the shape of the input data for a basic RNN layer in Keras/TensorFlow?

A. (Batch Size, Timesteps, Features)
B. (Batch Size, Features)
C. (Features, Labels)
D. (Timesteps, Batch Size)

28 Why are Bidirectional RNNs (BiRNNs) useful?

A. They eliminate the need for backpropagation
B. They allow the network to have context from both the past and the future
C. They train faster than standard RNNs
D. They use fewer parameters

29 In a Many-to-One sequence model (e.g., Sentiment Analysis), where is the output typically taken?

A. At every time step
B. Randomly sampled
C. At the last time step
D. At the first time step

30 What does the 'candidate cell state' in an LSTM do?

A. It proposes new values that could be added to the state
B. It clears the memory
C. It decides what to forget
D. It outputs the final prediction

31 Which issue leads to the 'Exploding Gradient' problem?

A. Learning rate being too low
B. Weights initialized to zero
C. Gradients > 1 accumulating multiplicatively
D. Gradients < 1 accumulating multiplicatively

32 A solution to the Exploding Gradient problem is:

A. Increasing the learning rate
B. Removing the hidden layer
C. Using ReLU
D. Gradient Clipping

33 In an Attention model, the vector c_t is often referred to as:

A. The Context Vector
B. The Bias Vector
C. The Forget Vector
D. The Noise Vector

34 Sequence-to-Sequence models are most commonly associated with:

A. Linear Regression
B. Cluster Analysis
C. Text Summarization
D. Image Segmentation

35 What is 'Global Attention'?

A. Attention that considers only a window of hidden states
B. Attention applied without weights
C. Attention applied to a single word
D. Attention that considers all hidden states of the encoder

36 Which of the following describes 'Greedy Decoding'?

A. Choosing the word with the highest probability at each step immediately
B. Considering all possible future sequences
C. Waiting until the end to choose words
D. Choosing a random word based on distribution

37 In POS tagging, if a word is ambiguous (e.g., 'book' can be a noun or verb), how does an RNN resolve it?

A. It flips a coin
B. It uses the context provided by surrounding words stored in the hidden state
C. It cannot resolve ambiguity
D. It always picks the most common usage

38 What is the typical loss function for a multi-class classification problem like POS Tagging or NMT?

A. Hinge Loss
B. Categorical Cross-Entropy
C. Mean Squared Error (MSE)
D. Absolute Error

39 What does GRU stand for?

A. General Regression Unit
B. Global Recurrent Update
C. Gated Recurrent Unit
D. Gradient Rectified Unit

40 The 'Input Gate' in an LSTM is usually controlled by which activation function?

A. Sigmoid
B. ReLU
C. Tanh
D. Linear

41 What visual tool is often used to interpret what an Attention model has learned?

A. Pie Chart
B. Scatter Plot
C. Histogram
D. Attention Heatmap

42 Which limitation of RNNs prevents parallelization during training?

A. Complex loss functions
B. Use of sigmoid functions
C. Sequential dependency of the hidden state
D. Large memory footprint

43 In a sequence model, 'padding' is used to:

A. Increase the learning rate
B. Remove stopwords
C. Make all sequences in a batch the same length
D. Add noise to the data

44 Which of these is a 'many-to-one' application of sequence models?

A. Machine Translation
B. Music Generation
C. Sentiment Classification
D. Video Captioning

45 In an NER task, identifying 'Apple' as an Organization rather than a Fruit relies on:

A. The length of the word
B. The contextual information in the sequence
C. The spelling of the word
D. The capitalization

46 Why is the traditional Encoder-Decoder model often described as having 'amnesia'?

A. It cannot learn new words
B. It uses a forget gate
C. It forgets the weights after training
D. It struggles to retain information from the beginning of a long sequence at the decoding stage

47 In the attention equation score(h_t, h_s), what are h_t and h_s?

A. Learning rate and Loss
B. Weight and Bias
C. Input and Output Gates
D. Decoder hidden state and Encoder hidden state

48 Which mechanism allows a model to focus on 'local' parts of the input sequence based on the current decoding step?

A. Dropout
B. Attention Mechanism
C. Max Pooling
D. Batch Normalization

49 In sequence labeling, what does the output layer usually consist of?

A. A clustering algorithm
B. A Softmax layer over the tag set for each time step
C. A single neuron
D. A linear regression layer

50 Which of the following best describes the 'Seq2Seq' mapping?

A. Variable Input Size -> Variable Output Size
B. Fixed Input Size -> Variable Output Size
C. Fixed Input Size -> Fixed Output Size
D. Variable Input Size -> Fixed Output Size