AI0-001 AI Concepts and Techniques Practice Question
A developer is using a pre-trained BERT model for a question-answering system. They want to ensure the model can handle out-of-vocabulary words. Which component of the BERT architecture is responsible for this?
⚠ Common exam trap
The trap here is that candidates often associate 'handling unknown words' with the attention mechanism or positional encoding, but the CompTIA exam specifically tests the understanding that tokenisation—not the model's internal layers—is what makes BERT robust to OOV words.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
WordPiece tokenisation
WordPiece tokenisation is the component of BERT that handles out-of-vocabulary (OOV) words by breaking them into subword units (e.g., 'playing' → 'play' + '##ing'). This allows the model to represent any word, even unseen ones, as a sequence of known subword tokens, ensuring no word is truly out of vocabulary.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Positional encoding
Why it's wrong here
Positional encoding injects token-order information into embeddings; it does not segment or represent unseen words. It is tempting because it operates on the input embeddings where vocabulary is handled, and it would be correct if the question asked how BERT preserves word order without recurrence.
- ✗
Feed-forward layers
Why it's wrong here
Feed-forward layers apply position-wise transformations to each token vector; they hold no vocabulary or subword segmentation, so they cannot resolve unknown words. They are tempting because they process every token, and they would be the answer if the question asked which component adds non-linear transformation capacity.
- ✓
WordPiece tokenisation
Why this is correct
WordPiece tokenisation splits unknown or rare words into frequently occurring subword units drawn from a fixed vocabulary, so the model represents out-of-vocabulary words as sequences of known subwords rather than a single unknown token. This is the component handling OOV input.
- ✗
Attention mechanism
Why it's wrong here
The attention mechanism computes weighted relationships between existing token representations; it cannot map an unseen word to a subword unit. It is tempting because it is BERT's defining component, and it would be correct if the question asked which component captures contextual dependencies between tokens.
About these practice questions
This AI0-001 question is part of Courseiva's 962-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI0-001 practice question is part of Courseiva's free CompTIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI0-001 exam.