NLP - Sequence Models for NLP
Why does the embedding layer output have 3 dimensions for input shape (batch_size, sequence_length)?
Why does the embedding layer output have 3 dimensions for input shape (batch_size, sequence_length)?
hard📝 Conceptual Q10 of 15
Step-by-Step Solution
Solution:
Step 1: Analyze input and output dimensions
Input has 2 dimensions: batch size and sequence length. Embedding converts each token to a vector, adding a third dimension.Step 2: Check other options
Embedding does not add batch dimension, does not convert to one-hot, and does not flatten sequences.Final Answer:
Because each token is converted to a vector, adding the embedding dimension -> Option AQuick Check:
Embedding output dims = input dims + embedding vector size [OK]
Quick Trick: Embedding adds vector dimension per token, increasing output dims [OK]
Common Mistakes:
MISTAKES- Thinking embedding adds batch dimension
- Confusing embedding with one-hot encoding
- Assuming embedding flattens input
Master "Sequence Models for NLP" in NLP
Want More Practice?
15+ quiz questions · All difficulty levels · Free
Free Signup - Practice All QuestionsMore NLP Quizzes