Natural Language Inference on Contract-NLI (test)
0.86macro-F1Post Conflict resolution (PLD)
Evaluation Results
| Method | Links | |
|---|---|---|
| Post Conflict resolution (PLD)Model=Gemini 3 Flash (Teacher)2026.02 | 0.86 | |
| Clustered Instructions (PLD)Model=Gemini 3 Flash (Teacher)2026.02 | 0.84 | |
| Post Conflict resolution (PLD)Model=Gemma2026.02 | 0.83 | |
| Post Conflict resolution (PLD)Model=Gemini 2 Flash2026.02 | 0.83 | |
| Clustered Instructions (PLD)Model=Gemma2026.02 | 0.81 | |
| Clustered Instructions (PLD)Model=Gemini 2 Flash2026.02 | 0.8 | |
| Few-shotModel=Gemini 3 Flash (Teacher)2026.02 | 0.79 | |
| Zero-shotModel=Gemini 3 Flash (Teacher)2026.02 | 0.77 | |
| Few-shotModel=Gemini 2 Flash2026.02 | 0.75 | |
| Zero-shotModel=Gemini 2 Flash2026.02 | 0.73 | |
| Few-shotModel=Gemma2026.02 | 0.7 | |
| Zero-shotModel=Gemma2026.02 | 0.67 |