Video Understanding on MLVU (Accuracy Delta Reporting)
68.4Base AccuracyAKS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| AKSModel=InternVL-3 [47], Size=8B, Frame=322026.02 | 68.4 | 74.2 | 5.8 | |
| MDP3Model=InternVL-3 [47], Size=8B, Frame=322026.02 | 68.4 | 74 | 5.6 | |
| A.I.R.Model=InternVL-3 [47], Size=8B, Frame=≤32§2026.02 | 68.4 | 74.5 | 6.1 | |
| WFS-SBModel=InternVL-3 [47], Size=8B, Frame=322026.02 | 68.4 | 74.8 | 6.4 | |
| KFCModel=LLaVA-Video [44], Size=7B, Frame=82026.02 | 60.5 | 66.9 | 6.4 | |
| FrameOracleModel=LLaVA-Video [44], Size=7B, Frame=82026.02 | 60.5 | 63.4 | 2.9 | |
| AKSModel=Qwen2.5-VL [2], Size=7B, Frame=322026.02 | 59.7 | 67.2 | 7.5 | |
| WFS-SBModel=Qwen2.5-VL [2], Size=7B, Frame=322026.02 | 59.7 | 70.4 | 10.7 | |
| MDP3Model=Qwen2.5-VL [2], Size=7B, Frame=322026.02 | 59.3 | 66.2 | 6.9 | |
| A.I.R.Model=Qwen2.5-VL [2], Size=7B, Frame=≤32§2026.02 | 59.3 | 67.5 | 8.2 | |
| BOLTModel=LLaVA-OV [10], Size=7B, Frame=82026.02 | 58.9 | 63.4 | 4.5 | |
| AKSModel=LLaVA-OV [10], Size=7B, Frame=82026.02 | 58.6 | 62.9 | 4.3 | |
| WFS-SBModel=LLaVA-OV [10], Size=7B, Frame=82026.02 | 58.6 | 67.2 | 8.6 | |
| Frame-VoyagerModel=LLaVA-OV [10], Size=7B, Frame=82026.02 | 58.5 | 65.6 | 7.1 | |
| KFCModel=LLaVA-OV [10], Size=7B, Frame=82026.02 | 58.5 | 66.2 | 7.7 | |
| FrameOracleModel=LLaVA-OV [10], Size=7B, Frame=82026.02 | 58.4 | 62.9 | 4.5 | |
| AKSModel=LLaVA-Video [44], Size=7B, Frame=82026.02 | 57.4 | 64.2 | 6.8 | |
| WFS-SBModel=LLaVA-Video [44], Size=7B, Frame=82026.02 | 57.4 | 66.9 | 9.5 |