Paired LLM Evaluation Resolution Analysis on MMLU-Pro headline
4Unresolved Pair CountFixed-n
Evaluation Results
| Method | Links | |
|---|---|---|
| Fixed-nProcedure=Fixed-n2026.05 | 4 | |
| Bonferroni / HolmProcedure=Bonferroni / Holm2026.05 | 4 | |
| Anytime-validProcedure=Anytime-valid2026.05 | 5 | |
| Real-ICC clusteredProcedure=Real-ICC clustered2026.05 | 6 |