Multimodal Understanding on MME^P
1,487.2ScoreUpper Bound
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Upper BoundTokens=5762025.12 | 1,487.2 | 100 | |
| HTC-VLMTokens=1 (hybrid)2025.12 | 1,265.2 | 74.5 | |
| VoCo-LLaMATokens=12025.12 | 1,241.4 | 71.7 | |
| Avg. PoolTokens=12025.12 | 1,210.3 | 81 | |
| Q-FormerTokens=12025.12 | 1,079.7 | 75.2 | |
| Lower BoundTokens=12025.12 | 617.3 | 0 |