Question Answering on HotpotQA single-predictor (val)
64Exact MatchTree contrastive
Evaluation Results
| Method | Links | |
|---|---|---|
| Tree contrastiveValidation slice=500-example2026.06 | 64 | |
| MIPROv2-lightMode=instruction-only2026.06 | 62.2 | |
| GEPA-lightOptimizer validation slice=100-example2026.06 | 61.2 | |
| Baseline promptSettings=instruction-only2026.06 | 56 |