Accuracy and TFLOPs on MedQA
64.57AccuracyOurs-context-unaware
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Ours-context-unawarePair=8B → 4B, Context Setting=context-unaware2026.06 | 64.57 | 6.56 | |
| Ours-context-unawarePair=4B → 8B, Context Setting=context-unaware2026.06 | 61.82 | 9.42 | |
| Ours-context-unawarePair=14B → 4B, Context Setting=context-unaware2026.06 | 59.54 | 8.9 | |
| Ours-context-unawarePair=8B → 14B, Context Setting=context-unaware2026.06 | 57.74 | 15.85 | |
| Ours-context-unawarePair=14B → 8B, Context Setting=context-unaware2026.06 | 54.2 | 13.28 | |
| Ours-context-unawarePair=4B → 14B, Context Setting=context-unaware2026.06 | 53.26 | 17.15 | |
| T2T-context-unawarePair=8B → 14B, Context Setting=context-unaware2026.06 | 27.81 | 72.06 | |
| T2T-context-unawarePair=4B → 14B, Context Setting=context-unaware2026.06 | 27.65 | 60.83 | |
| T2T-context-unawarePair=14B → 8B, Context Setting=context-unaware2026.06 | 25.22 | 53.8 | |
| T2T-context-unawarePair=8B → 4B, Context Setting=context-unaware2026.06 | 22.94 | 32.75 | |
| T2T-context-unawarePair=14B → 4B, Context Setting=context-unaware2026.06 | 21.13 | 41.95 | |
| T2T-context-unawarePair=4B → 8B, Context Setting=context-unaware2026.06 | 18.85 | 37.86 | |
| C2C-context-unawarePair=4B → 14B, Context Setting=context-unaware2026.06 | 12.8 | 20.05 | |
| C2C-context-unawarePair=4B → 8B, Context Setting=context-unaware2026.06 | 8.96 | 14.94 | |
| C2C-context-unawarePair=8B → 4B, Context Setting=context-unaware2026.06 | 7.23 | 14.76 | |
| C2C-context-unawarePair=8B → 14B, Context Setting=context-unaware2026.06 | 6.36 | 9.84 | |
| C2C-context-unawarePair=14B → 8B, Context Setting=context-unaware2026.06 | 5.18 | 24.77 | |
| C2C-context-unawarePair=14B → 4B, Context Setting=context-unaware2026.06 | 0.63 | 17.93 |