Logical Reasoning on MNLogic 7dgt
60.31AccuracyVLC
Evaluation results
| Method | Links | |
|---|---|---|
| VLCBackbone Scale=7B, Random Seeds=52026.03 | 60.31 | |
| End2end FTBackbone Scale=7B, Random Seeds=52026.03 | 51.95 | |
| PrismBackbone Scale=7B, Random Seeds=52026.03 | 50.53 | |
| End2end RSBackbone Scale=7B, Random Seeds=52026.03 | 50.05 | |
| ViperGPTBackbone Scale=7B, Random Seeds=52026.03 | 42.74 |