AI0-001 AI Concepts and Foundations Practice Question
A team is training a deep learning model for natural language processing using a large corpus. They notice the model has a very high number of parameters and training is slow. Which technique can reduce the number of parameters without significant performance loss?
⚠ Common exam trap
Test-takers frequently confuse regularization techniques (like dropout) or training speed optimizations (batch size, learning rate) with actual parameter reduction, which only embedding compression directly achieves.
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
✓
Apply embedding compression
Embedding compression reduces the dimensionality of the embedding layer, which often contains the majority of the model's parameters in NLP tasks. By using techniques like low-rank factorization or pruning, the model retains most of its representational power while significantly decreasing the parameter count and training time.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Apply embedding compression
Why this is correct
Embedding compression reduces the dimensionality or parameter count of embedding layers through techniques such as low-rank factorisation or weight sharing, directly cutting total parameters. This addresses the slow training caused by the oversized embedding table while retaining semantic representation quality.
- ✗
Add more dropout layers
Why it's wrong here
Dropout regularises activations during training but leaves the parameter count unchanged at inference. It is tempting because it combats overfitting on large corpora, and it would be correct if the issue were generalisation; the stem instead asks for fewer parameters.
- ✗
Use a larger batch size
Why it's wrong here
Larger batches change gradient averaging and hardware utilisation, not parameter count; the weight matrices stay identical. It is tempting because bigger batches speed up each epoch, which is the correct fix when training throughput, not model size, is the bottleneck.
- ✗
Increase learning rate
Why it's wrong here
Raising the learning rate alters step size along the loss surface; it cannot remove weights or shrink layers. It is tempting because a higher rate can accelerate convergence, which is the right lever when training is slow purely due to cautious optimisation rather than parameter volume.
About these practice questions
Courseiva writes every AI0-001 question from scratch — 962 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or 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.