EvalFloor: is your LLM eval improvement real? Trying k prompt variants and keeping the best scores points on noise alone — this computes how many.
-
Updated
Sep 21, 2026 - Python
#
winners-curse
Here are 3 public repositories matching this topic...
Closed Q1 2026: Kalshi weather-contracts trading system. Calibration validation failed at Gate 1; full post-mortem included. Methodology extracted to ai-discipline.
-
Updated
May 22, 2026 - Python
Your A/B test winner is about half as good as it looked. The winner's curse, peeking and power measured on 32,363 real randomised experiments -- with a live demo that re-splits the traffic in your browser.
python data-science statistics experimentation ab-testing causal-inference empirical-bayes fastapi sequential-testing winners-curse
-
Updated
Aug 12, 2026 - Python
Add this topic to your repo
To associate your repository with the winners-curse topic, visit your repo's landing page and select "manage topics."