Attention mechanism in depth in NLP - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is the main purpose of the attention mechanism in neural networks?

The attention mechanism helps the model focus on the most important parts of the input data when making predictions, similar to how humans pay attention to relevant information.

intermediate

Explain the difference between 'soft' and 'hard' attention.

Soft attention assigns weights to all input parts and computes a weighted sum, allowing smooth focus. Hard attention selects one part of the input, making it discrete and non-differentiable, often requiring special training methods.

beginner

What are the three main components of the scaled dot-product attention?

The three components are Query (Q), Key (K), and Value (V). The attention score is computed by comparing Q with K, then used to weight V for the output.

intermediate

Why do we scale the dot product by the square root of the key dimension in scaled dot-product attention?

Scaling by the square root of the key dimension prevents the dot product values from becoming too large, which can cause very small gradients and slow learning.

intermediate

How does multi-head attention improve the model's ability to focus on different parts of the input?

Multi-head attention runs several attention mechanisms in parallel, each focusing on different parts or aspects of the input, allowing the model to capture diverse information.

What does the 'Query' represent in the attention mechanism?

A The information used to compare with keys

B The part of the input we want to focus on

C The output of the attention layer

D The weights assigned to input tokens

The Query is used to compare with Keys to calculate attention scores.

Why is softmax used in attention mechanisms?

A To select the maximum value only

B To increase the size of the input

C To reduce the number of parameters

D To normalize attention scores into probabilities

Softmax converts raw attention scores into probabilities that sum to 1.

Which of these is NOT a benefit of multi-head attention?

A Captures information from different representation subspaces

B Allows the model to attend to multiple positions simultaneously

C Reduces the total number of parameters drastically

D Improves the model's ability to understand complex relationships

Multi-head attention increases parameters due to multiple heads, not reduces them.

What problem does the attention mechanism help solve in sequence models?

A Vanishing gradients in deep networks

B Difficulty in remembering long-range dependencies

C Overfitting on small datasets

D Reducing training time by skipping layers

Attention helps models remember and focus on important parts of long sequences.

In scaled dot-product attention, what happens after computing the dot product between Query and Key?

A The result is scaled and passed through softmax to get weights

B The result is multiplied by the Value directly

C The result is ignored and only Value is used

D The result is passed through a ReLU activation

The dot product is scaled and softmaxed to produce attention weights.

Describe how the attention mechanism works step-by-step in a neural network.

Think about how the model decides what to focus on.

Explain why multi-head attention is more powerful than single-head attention.

Imagine looking at a picture from different angles.

Practice

1. What is the main purpose of the attention mechanism in NLP models?

easy

A. To increase the size of the input data

B. To reduce the number of layers in the model

C. To help the model focus on important parts of the input data

D. To randomly shuffle the input tokens

Solution

1. Step 1: Understand attention's role

   Attention helps models decide which parts of the input are most important for the task.

2. Step 2: Compare options

   Only To help the model focus on important parts of the input data correctly describes this focus mechanism; others describe unrelated actions.

3. Final Answer:

   To help the model focus on important parts of the input data -> Option C

4. Quick Check:

   Attention = Focus on important input [OK]

Hint: Remember: attention means focusing on key input parts [OK]

Common Mistakes:

• Thinking attention changes input size
• Confusing attention with model depth
• Assuming attention shuffles data

2. Which of the following correctly represents the formula for attention weights using queries (Q), keys (K), and softmax?

easy

A. softmax(Q x K^T)

B. Q + K

C. softmax(Q - K)

D. Q x K

Solution

1. Step 1: Recall attention weight calculation

   Attention weights are computed by multiplying queries with keys transposed, then applying softmax.

2. Step 2: Evaluate options

   Only softmax(Q x K^T) matches the correct formula softmax(Q x K^T). Others are incorrect operations.

3. Final Answer:

   softmax(Q x K^T) -> Option A

4. Quick Check:

   Attention weights = softmax(Q x K^T) [OK]

Hint: Attention weights = softmax of query-key dot product [OK]

Common Mistakes:

• Using addition instead of multiplication
• Forgetting to transpose keys
• Skipping softmax normalization

3. Given queries Q = [[1, 0]], keys K = [[1, 0], [-10, 1]], and values V = [[10, 20], [30, 40]], what is the output of the attention mechanism (using dot product and softmax)?

medium

A. [[10, 20]]

B. [[20, 30]]

C. [[20, 40]]

D. [[30, 40]]

Solution

1. Step 1: Calculate dot products Q x K^T

   Q = [1,0], K = [[1,0],[-10,1]]; dot products: [1*1+0*0=1, 1*(-10)+0*1=-10]

2. Step 2: Apply softmax to scores

   softmax([1,-10]) ≈ [1, 0] (e^{-10} negligible)

3. Step 3: Compute weighted sum of values

   Output ≈ 1*[10,20] + 0*[30,40] = [[10, 20]]

4. Step 4: Match option

   [[10, 20]] matches exactly.

5. Final Answer:

   [[10, 20]] -> Option A

6. Quick Check:

   Weighted sum of values = [[10, 20]] [OK]

Hint: Calculate dot, softmax, then weighted sum of values [OK]

Common Mistakes:

• Skipping softmax normalization
• Using keys instead of values for output
• Incorrect dot product calculation

4. Identify the error in this attention weight calculation code snippet:

import numpy as np
Q = np.array([[1, 0]])
K = np.array([[1, 0], [-10, 1]])
scores = np.dot(Q, K)
weights = np.exp(scores) / np.sum(np.exp(scores))

medium

A. Values are missing in the calculation

B. Softmax is applied incorrectly

C. Queries and keys have incompatible shapes

D. Keys should be transposed before dot product

Solution

1. Step 1: Check dot product operation

   Dot product should be between Q and K transposed to align dimensions correctly.

2. Step 2: Analyze code

   Code uses np.dot(Q, K) without transposing K, causing wrong shape and incorrect scores.

3. Final Answer:

   Keys should be transposed before dot product -> Option D

4. Quick Check:

   Transpose keys before dot product [OK]

Hint: Always transpose keys before dot product with queries [OK]

Common Mistakes:

• Forgetting to transpose keys
• Misapplying softmax formula
• Ignoring shape compatibility

5. In a transformer model, why is scaling the dot product by the square root of the key dimension important before applying softmax?

hard

A. To increase the dot product values for better attention

B. To prevent large dot product values causing very small gradients

C. To normalize the values between 0 and 1

D. To reduce the number of keys used in attention

Solution

1. Step 1: Understand dot product scaling

   Large dot products can cause softmax to produce very small gradients, slowing learning.

2. Step 2: Role of scaling by sqrt of key dimension

   Scaling reduces dot product magnitude, stabilizing gradients and improving training.

3. Final Answer:

   To prevent large dot product values causing very small gradients -> Option B

4. Quick Check:

   Scaling avoids tiny gradients in softmax [OK]

Hint: Scale dot product to keep gradients healthy [OK]

Common Mistakes:

• Thinking scaling increases dot product
• Confusing scaling with normalization to [0,1]
• Assuming scaling reduces keys count