Self-attention and multi-head attention in NLP - Model Metrics & Evaluation

Metrics & Evaluation - Self-attention and multi-head attention

Which metric matters for Self-attention and Multi-head Attention and WHY

In models using self-attention and multi-head attention, like Transformers, the key metrics to check are accuracy or loss on the task (e.g., translation, text classification). These metrics show how well the model understands relationships in the input.

Since attention helps the model focus on important parts of the input, improvements in accuracy or reduction in loss indicate better attention learning.

For sequence tasks, metrics like BLEU (for translation) or F1-score (for classification) are also important to measure quality.

Confusion Matrix Example for Attention-based Text Classification

      | Predicted Positive | Predicted Negative |
      |--------------------|--------------------|
      | True Positive (TP): 85 | False Negative (FN): 15 |
      | False Positive (FP): 10 | True Negative (TN): 90 |
    

Total samples = 85 + 15 + 10 + 90 = 200

Precision = TP / (TP + FP) = 85 / (85 + 10) = 0.894

Recall = TP / (TP + FN) = 85 / (85 + 15) = 0.85

F1-score = 2 * (Precision * Recall) / (Precision + Recall) ≈ 0.871

Precision vs Recall Tradeoff in Attention Models

Attention models can be tuned to focus more on precision or recall depending on the task.

• High Precision: The model is very sure about its positive predictions. Useful when false alarms are costly, like spam detection.
• High Recall: The model finds most positive cases, even if some are wrong. Important in medical diagnosis to catch all cases.

Multi-head attention helps by looking at input from different views, improving both precision and recall by capturing diverse information.

Good vs Bad Metric Values for Attention Models

Good: Accuracy above 85%, F1-score above 0.85, balanced precision and recall showing the model understands input relations well.

Bad: Accuracy near random chance (e.g., 50% for binary), very low recall or precision (below 0.5), indicating the attention mechanism is not helping the model focus correctly.

Common Pitfalls in Metrics for Attention Models

• Accuracy Paradox: High accuracy but poor recall or precision can mislead about model quality.
• Data Leakage: If training data leaks into test, metrics look better but model won't generalize.
• Overfitting: Very low training loss but high test loss means attention learned noise, not true patterns.
• Ignoring Class Imbalance: Metrics like accuracy can be misleading if classes are uneven; use F1 or AUC instead.

Self-Check Question

Your Transformer model with multi-head attention has 98% accuracy but only 12% recall on the positive class (e.g., fraud). Is it good for production?

Answer: No, it is not good. The low recall means the model misses most positive cases, which is critical in fraud detection. Despite high accuracy, the model fails to catch fraud effectively.

Key Result

Precision, recall, and F1-score are key to evaluate how well self-attention and multi-head attention models focus on important input parts and balance correct predictions.

Practice

1. What is the main purpose of self-attention in natural language processing?

easy

A. To reduce the size of the input data by removing words

B. To generate random sentences without context

C. To translate text from one language to another

D. To let the model focus on important words by comparing all words to each other

Solution

1. Step 1: Understand self-attention's role

   Self-attention helps the model look at all words in a sentence and decide which ones are important by comparing them to each other.

2. Step 2: Match purpose with options

   To let the model focus on important words by comparing all words to each other correctly describes this focus mechanism, while others describe unrelated tasks.

3. Final Answer:

   To let the model focus on important words by comparing all words to each other -> Option D

4. Quick Check:

   Self-attention = focus on important words [OK]

Hint: Self-attention means comparing words to find importance [OK]

Common Mistakes:

• Confusing self-attention with translation
• Thinking self-attention removes words
• Assuming it generates random text

2. Which of the following is the correct way to describe multi-head attention?

easy

A. Running several self-attention processes in parallel to get richer understanding

B. Applying self-attention only once on the input

C. Using attention only on the first word of a sentence

D. Ignoring word relationships and focusing on word order only

Solution

1. Step 1: Recall multi-head attention definition

   Multi-head attention means running multiple self-attention operations at the same time to capture different aspects of word relationships.

2. Step 2: Compare options to definition

   Running several self-attention processes in parallel to get richer understanding matches this exactly; others describe incomplete or incorrect ideas.

3. Final Answer:

   Running several self-attention processes in parallel to get richer understanding -> Option A

4. Quick Check:

   Multi-head attention = multiple self-attentions [OK]

Hint: Multi-head means many self-attentions at once [OK]

Common Mistakes:

• Thinking multi-head means single attention
• Believing it focuses only on first word
• Ignoring word relationships

3. Given the following simplified self-attention scores matrix for a 3-word sentence:
Scores = [[1, 0.5, 0], [0.5, 1, 0.2], [0, 0.2, 1]]
What is the attention weight for the second word attending to the third word after applying softmax on its row?

medium

A. Approximately 0.21

B. Approximately 0.50

C. Approximately 0.29

D. Approximately 0.70

Solution

1. Step 1: Extract the second row scores

   The second word's scores are [0.5, 1, 0.2].

2. Step 2: Apply softmax to these scores

   Softmax formula: exp(score) / sum(exp(all scores)). Calculate exp(0.5)=1.65, exp(1)=2.72, exp(0.2)=1.22. Sum = 1.65+2.72+1.22=5.59. Attention weight for third word = 1.22/5.59 ≈ 0.218.

3. Final Answer:

   Approximately 0.21 -> Option A

4. Quick Check:

   Softmax normalizes scores to probabilities [OK]

Hint: Softmax turns scores into probabilities summing to 1 [OK]

Common Mistakes:

• Forgetting to exponentiate scores
• Dividing by wrong sum
• Mixing row and column values

4. Consider this Python code snippet for multi-head attention weights calculation:

import numpy as np

def multi_head_attention(scores_list):
    heads = []
    for scores in scores_list:
        weights = np.exp(scores) / np.sum(np.exp(scores))
        heads.append(weights)
    return np.mean(heads, axis=0)

scores_list = [np.array([1, 0, 2]), np.array([0, 1, 1])]
print(multi_head_attention(scores_list))

What is the main bug in this code?

medium

A. Softmax is applied incorrectly; denominator should sum over exp(scores) per head

B. The function returns mean of weights instead of concatenating heads

C. The code uses np.exp twice causing overflow

D. Scores_list should be a 2D array, not a list of arrays

Solution

1. Step 1: Analyze softmax calculation

   Softmax is correctly applied per head by dividing exp(scores) by sum of exp(scores).

2. Step 2: Check output aggregation

   The function averages the weights from each head, but multi-head attention should concatenate or combine heads differently, not average weights element-wise.

3. Final Answer:

   The function returns mean of weights instead of concatenating heads -> Option B

4. Quick Check:

   Multi-head attention combines heads, not averages weights [OK]

Hint: Multi-head attention concatenates heads, not averages weights [OK]

Common Mistakes:

• Thinking averaging weights is correct
• Confusing softmax denominator
• Assuming input format is wrong

5. You want to improve a Transformer model's ability to understand complex sentences by increasing the number of attention heads from 4 to 8. What is the most likely effect of this change?

hard

A. The model will ignore word order completely

B. The model will run faster but lose accuracy

C. The model can capture more diverse word relationships but may require more computation

D. The model will only focus on the first half of the sentence

Solution

1. Step 1: Understand effect of increasing attention heads

   More heads mean the model can look at different parts of the sentence simultaneously, capturing richer relationships.

2. Step 2: Consider computational cost and accuracy

   Increasing heads usually increases computation and memory needs but can improve understanding and accuracy.

3. Final Answer:

   The model can capture more diverse word relationships but may require more computation -> Option C

4. Quick Check:

   More heads = richer focus + more compute [OK]

Hint: More heads = better focus but slower model [OK]

Common Mistakes:

• Assuming more heads always make model faster
• Thinking word order is ignored
• Believing model focuses only on part of sentence