Logical Reasoning on FOLIO v0.0 (val)
44.1AccuracyPhi-2
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Phi-2n=2042026.06 | 44.1 | 41.7 | 56.5 | 1.164 | 19.5 | 78.9 | |
| Qwen2.5-1.5Bn=2042026.06 | 40.2 | 30.9 | 50.8 | 0.674 | 16.6 | 46.1 | |
| Qwen2.5-3Bn=2042026.06 | 38.2 | 7.4 | 80 | 0.115 | 2.5 | 6.4 | |
| TinyLlama-1.1Bn=2042026.06 | 34.3 | 79.4 | 34.6 | 1.698 | 69.8 | 100 |