AI0-001 AI Concepts and Foundations Practice Question
A hospital's AI governance committee is reviewing a sepsis-prediction model before deployment. The model was trained on five years of historical ICU data in which patients who received early antibiotics had better outcomes, and the model learned to recommend antibiotics for nearly every patient with any fever. The committee wants to determine whether the model has learned a spurious correlation rather than a true clinical signal. Which action best evaluates this concern?
⚠ Common exam trap
The trap here is assuming that a high AUC or strong test-set performance proves the model learned a genuine clinical relationship, when a spurious correlation present throughout the data can produce equally strong metrics.
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
✓
Perform a feature-ablation study, removing fever and antibiotic-administration variables, and measure the change in predictive performance and decision patterns.
Feature ablation is the most direct way to test whether a model depends on suspected spurious inputs. By removing fever and antibiotic-administration variables and observing changes in performance and recommendation behavior, the committee can determine whether the model's decisions hinge on those confounded features rather than on genuine clinical indicators of sepsis.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Increase the size of the test set by repartitioning the existing data and recompute the AUC.
Why it's wrong here
Repartitioning changes only how the existing data is divided, not what the model learned. A larger test set can reduce variance in the AUC estimate, but if the spurious correlation exists throughout the dataset, the AUC will remain high and misleading. This action cannot distinguish a genuine clinical signal from a confounded one because it never intervenes on the model's inputs.
- ✗
Deploy the model in shadow mode and compare its alerts against clinician judgment for one month.
Why it's wrong here
Shadow deployment compares model outputs to human decisions, but clinicians in the historical setting also acted under the same confounding pattern, so agreement does not prove the model learned a true signal. It is useful for workflow validation and safety monitoring, yet it cannot isolate whether fever is functioning as a spurious proxy. The evaluation would be confounded by the very behavior under investigation.
- ✓
Perform a feature-ablation study, removing fever and antibiotic-administration variables, and measure the change in predictive performance and decision patterns.
Why this is correct
Ablation directly tests whether the model's recommendations depend on the suspected spurious features. If removing fever and antibiotic variables sharply degrades performance or changes which patients receive recommendations, the model likely relied on the confound. This is a targeted causal-probing technique that reveals reliance on specific inputs rather than overall accuracy, which is exactly what the committee needs to assess.
- ✗
Retrain the model with a larger learning rate and compare training loss across epochs.
Why it's wrong here
Learning rate controls how quickly weights update during optimization; it does not test whether the model's features reflect causal clinical relationships. A larger learning rate may speed convergence or cause instability, but comparing training loss across epochs only reveals optimization behavior, not whether fever is being used as a spurious proxy for antibiotic response. This action leaves the suspected confound entirely unexamined.
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 and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official CompTIA exam blueprint
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.