Reasoning on BBH-Hard 2022
22.9Peak AccuracyPessimism Only
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Pessimism Only2026.04 | 22.9 | 22.1 | 0.9 | |
| RM + Pessimism2026.04 | 18.5 | 11 | 7.5 | |
| Reward Model2026.04 | 17.3 | 1.7 | 15.6 |
| Method | Links | |||
|---|---|---|---|---|
| Pessimism Only2026.04 | 22.9 | 22.1 | 0.9 | |
| RM + Pessimism2026.04 | 18.5 | 11 | 7.5 | |
| Reward Model2026.04 | 17.3 | 1.7 | 15.6 |