Module 07 Knowledge Check
5 questions. Pass mark 4 out of 5. Answer every question before checking the answer key below, then retry after reading the feedback.
1. Meaningful human review requires that the reviewer can…
- A. See a summary
- B. Inspect the proposed action and its evidence, then approve, edit, or refuse before it executes
- C. Be notified afterwards
- D. Read the logs
2. An interrupt should be placed…
- A. At the end of the workflow
- B. Immediately before the consequential action
- C. At the start
- D. After the action, for audit
3. Letting a reviewer edit the proposed action means the workflow must…
- A. Discard the edit
- B. Resume using the edited state, not the original proposal
- C. Restart from scratch
- D. Ask the model again
4. Approval fatigue is a real failure mode because…
- A. It slows things
- B. Reviewers rubber-stamp everything when asked too often, so the control stops working
- C. It costs money
- D. It confuses the model
5. The review UI should surface…
- A. The raw graph
- B. What will happen, to whom, based on what evidence, and how to refuse
- C. Token counts
- D. The model name
---
Answer Key and Explanations
Check these only after attempting every question.
1. B - Inspect the proposed action and its evidence, then approve, edit, or refuse before it executes
Approval that cannot prevent execution is theatre, not control.
2. B - Immediately before the consequential action
The pause must sit at the point where refusal still changes the outcome.
3. B - Resume using the edited state, not the original proposal
If the edit is not carried into state, the human intervention is silently ignored.
4. B - Reviewers rubber-stamp everything when asked too often, so the control stops working
Gate the consequential steps, not every step, or the gate becomes a reflex click.
5. B - What will happen, to whom, based on what evidence, and how to refuse
A reviewer can only be accountable for a decision they were given enough context to make.
Practical Check
Apply this module to your own work: complete the module activity for *Human-in-the-Loop Review and Edits*, then write one sentence naming what your result shows and one naming what it does not.
Strong Answer Pattern
A strong answer names the task, the evidence used, the check performed, and the remaining limitation. It avoids "proved", "guaranteed", or "always" unless the evidence genuinely supports it.