Evaluations and datasets · Lesson 6 of 18 · 8 min

Read aggregate scores

Aggregate scores compress many case outcomes. Inspect the denominator, metric direction and important slices before treating one number as a release verdict.

Open the lab Course map

Mastery 0/3Run itExplain itApply it

learner@clairos:~ $

↑ history · Tab complete · Ctrl+C cancel · Esc, Tab to leaveExit 0

Your objectives

Hints are always available. Wrong turns never cost XP.

1
Read the case evidence
Read /opt/ai-lab/cases/aiops-evals-datasets-006.json. Separate the recorded facts from your assumptions.
Hint & worked example
Aggregate scores compress many case outcomes. Inspect the denominator, metric direction and important slices before treating one number as a release verdict.
The complete case JSON is read successfully.
```
cat /opt/ai-lab/cases/aiops-evals-datasets-006.json
```
Common mistake to avoid: A high aggregate guarantees good performance for every user group.
2
Isolate the deciding signal
Select the critical_slice_pass_rate field from the case JSON. What does it establish—and what does it not establish?
Hint & worked example
Inspect the weak critical slice despite a 92 percent aggregate pass rate.
A successful JSON query isolates critical_slice_pass_rate.
```
jq .critical_slice_pass_rate /opt/ai-lab/cases/aiops-evals-datasets-006.json
```
Common mistake to avoid: Using a different case or an unrelated signal as the justification.
3
Record your operational decision
Choose the justified code: inspect-critical-slice, change-without-evidence, or ignore-the-signal. Save that one code followed by a newline in /home/learner/lab/aiops-evals-datasets-006-decision.txt.
Hint & worked example
Inspect the weak critical slice despite a 92 percent aggregate pass rate.
The decision artifact contains the correct code, not just a matching command in history.
```
printf '%s\n' 'inspect-critical-slice' > /home/learner/lab/aiops-evals-datasets-006-decision.txt
```
Common mistake to avoid: A high aggregate guarantees good performance for every user group.

Understand the pattern

Aggregate scores compress many case outcomes. Inspect the denominator, metric direction and important slices before treating one number as a release verdict.

Interpret the evidence

Inspect the weak critical slice despite a 92 percent aggregate pass rate.

A tempting mistake

A high aggregate guarantees good performance for every user group.

All names, measurements, prices, versions, and thresholds in this exercise are fictional training data. The evidence supports this bounded scenario, not a claim about a live provider.

Worked example and expected result

$ cat /opt/ai-lab/cases/aiops-evals-datasets-006.json

{
  "fixture": "Fictional training data; not a live measurement.",
  "passed": 92,
  "evaluated": 100,
  "aggregate_pass_rate": 0.92,
  "critical_slice_pass_rate": 0.6
}

Exit status: 0. The command completed successfully in the training fixture.

$ jq .critical_slice_pass_rate /opt/ai-lab/cases/aiops-evals-datasets-006.json

0.6

Exit status: 0. The command completed successfully in the training fixture.

$ printf '%s\n' 'inspect-critical-slice' > /home/learner/lab/aiops-evals-datasets-006-decision.txt

(no output; exit 0)

Exit status: 0. The command completed successfully in the training fixture.

Try a variation

Name the assumption that would have to change for “inspect-critical-slice” to stop being the justified action. Do not silently replace missing evidence with a guess.

Explain it

What does the evidence in “Read aggregate scores” justify?A high aggregate guarantees good performance for every user group.Make the change now and gather the missing evidence afterward.Inspect the weak critical slice despite a 92 percent aggregate pass rate.

Apply it elsewhere

A colleague concludes: “A high aggregate guarantees good performance for every user group.” What should you do?Treat this fixture as a live measurement.Reject that conclusion; return to the case evidence and its stated limits.Accept it because the training command completed.

Keep the momentum.

Complete the lab, then explain and apply the pattern.

Next: Inspect failed cases Try an incident drill

Anonymous lesson and mastery progress lasts for this session. Sign in to keep synced lesson XP between sessions; mastery stays device-local. No speed bonus, lost lives, or hint penalties.

Keep learning

Case methodology: evidence, limits, decision