Skip to content

Topic Drill 6.1.2: ML Model Documentation and Review ​

About this drill

LO AI-6.1.2 (K2): Explain the purpose and focus of reviewing ML model documentation.
Lesson: Ch 6, section 6.1.2

This drill covers every point in section 6.1.2:

  • Why documentation is critical for ML models; the role of transparency
  • Standardized documentation, regulatory compliance (EU AI Act, audits for high-risk systems)
  • Model Cards and Datasheets for Datasets
  • The eight checklist categories; the five aims of documentation review

Topic Drill: 6.1.2

Q1. Why does documentation become the MAIN tool for understanding, evaluating and trusting ML models?

ABecause ML models are always small and simple
BBecause regulators read only documentation
CBecause machine-generated code has limited understandability, models are black boxes that depend on data, and they are frequently updated
DBecause documentation can be generated automatically

Q2. What role does transparency play in ML model documentation, according to the syllabus?

AIt allows stakeholders to trace model behavior, decision logic and data lineage throughout the AI lifecycle
BIt hides proprietary details from regulators
CIt replaces testing of the model
DIt speeds up model training

Q3. For high-risk AI-based systems under frameworks such as the EU AI Act, what does the syllabus say about documentation?

ADocumentation is optional
BDocumentation is only needed after deployment
COnly the source code must be documented
DPassing a documentation audit is often a prerequisite for deployment

Q4. Which documentation framework gives a concise overview of a MODEL's intended uses, evaluation results and ethical considerations?

ADatasheets for Datasets
BModel Cards
CISO/IEC 25059
DA test summary report

Q5. What does "Datasheets for Datasets" standardize?

AThe model's hyperparameters
BThe deployment plan
CThe description of datasets, including their motivation, composition, collection process and uses
DThe API of the model

Q6. In the documentation checklist, which category contains "deployment plan, monitoring metrics/alerts, retraining strategy, model update/rollback plan, deprecation plan"?

AOperational
BUsage
CDesign
DTesting

Q7. In the documentation checklist, which category contains "self-learning approach, thresholds, data drift, concept drift"?

AFunctional
BDatasets
CNon-functional
DUsage

Q8. Reviewing a model card, a tester finds no information on the test dataset used or on how independent the testing was. Which checklist category is incomplete?

AGeneral
BTesting
CDesign
DOperational

Q9. Which TWO are aims of reviewing ML model documentation against a checklist, according to the syllabus?

Select 2 options.

ARetraining the model with better data
BIdentifying missing information, inaccuracies and inconsistencies
CReplacing dynamic testing of the model
DGenerating additional training data
EVerifying that all relevant regulatory requirements have been met

0 / 9 answered

← 6.1.1 Model Risks · Next drill: 6.1.3 Statistical Performance Testing →

Unofficial study guide for the ISTQB® Certified Tester AI Testing (CT-AI) exam. Not affiliated with International Software Testing Qualifications Board (ISTQB®).