individual-fairness.md's claimed sample output doesn't match its own seeded code

[yakew7]

Where: explainers/individual-fairness.md, lines ~113-115: the comment following a np.random.seed(42), n=2000 code block claims:

# Matched pairs: 142
# Individual fairness violations: 31
# Violation rate: 21.8%

Repro: running that exact code (verified the numpy seed-42 stream matches the canonical values first, via np.random.randint(0,100,10) -> [51 92 14 71 ...], so this isn't a cross-platform artifact) gives:

Matched pairs: 44
Individual fairness violations: 11
Violation rate: 25.0%

Reproducible on repeated runs. The claimed output doesn't match what the code actually produces - matched-pair count is off by more than 3x.

Fix direction: regenerate the claimed output from an actual run of the code block, and update the comment to match.