aggregation-aware-calibration-for-depth-confounder
IN premise — summaries/2026/08/24/aristotelian-2026-s2-related-work.md
Created 2026-08-24T17:10:50+00:00
Aggregation-aware calibration computes the null distribution of the same aggregate statistic (e.g., max over all L_A × L_B layer pairs) that is reported, rather than calibrating individual layer-pair scores, to address the depth confounder's multiple-comparisons inflation.
Summary
When you test thousands of layer pairs and report only the best-looking result, the score is inflated just by luck of having many chances. The fix is to calibrate that final "best-of-many" number directly against what it would look like under pure noise, rather than calibrating each pair in isolation and hoping the aggregation works out.