Logical Reasoning on RP Npred ≤ 30, δ ≤ 6 (train)
100Accuracydirect w. L = 24
Evaluation results
| Method | Links | |
|---|---|---|
| direct w. L = 24Reasoning Strategy=Direct, Number of Layers=242026.05 | 100 | |
| direct w. L = 128Reasoning Strategy=Direct, Number of Layers=1282026.05 | 100 | |
| cot w. L = 64Reasoning Strategy=CoT, Number of Layers=642026.05 | 100 | |
| cot w. L = 128Reasoning Strategy=CoT, Number of Layers=1282026.05 | 100 | |
| cot (baseline)Reasoning Strategy=CoT2026.05 | 99.9 | |
| direct (baseline)Reasoning Strategy=Direct2026.05 | 98.4 |