Long-video Understanding on VideoNIAH
46.81AccuracyVideo-Zero
Evaluation Results
| Method | Links | |
|---|---|---|
| Video-ZeroBackbone=MiMo-SFT-7B, Checkpoint=v12026.05 | 46.81 | |
| Video-ZeroBackbone=MiMo-SFT-7B, Checkpoint=v22026.05 | 46.74 | |
| MiMo-SFT-7BBackbone=MiMo-SFT-7B, Checkpoint=Base2026.05 | 46.67 | |
| Video-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=42026.05 | 46.37 | |
| Video-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=42026.05 | 46.37 | |
| Video-Zero (Iter 4)Backbone=Qwen3-VL-4B-Instruct2026.05 | 46.37 | |
| Video-Zero (Peak)Backbone=Qwen3-VL-4B-Instruct2026.05 | 46.37 | |
| Video-Zero (Iter 4)Backbone=Qwen3-VL-8B-Instruct2026.05 | 46.37 | |
| Video-Zero (Peak)Backbone=Qwen3-VL-8B-Instruct2026.05 | 46.37 | |
| Video-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=52026.05 | 46.15 | |
| Video-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=52026.05 | 46.07 | |
| V-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=42026.05 | 46.07 | |
| Video-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=22026.05 | 46.07 | |
| V-ZeroBackbone=Qwen3-VL-8B-Instruct2026.05 | 46.07 | |
| Video-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=32026.05 | 45.93 | |
| VisPlayBackbone=Qwen3-VL-8B-Instruct, Iteration=12026.05 | 45.85 | |
| VisPlayBackbone=Qwen3-VL-8B-Instruct, Iteration=52026.05 | 45.85 | |
| Video-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=12026.05 | 45.85 | |
| Base ModelBackbone=Qwen3-VL-4B-Instruct2026.05 | 45.78 | |
| VisPlayBackbone=Qwen3-VL-4B-Instruct, Iteration=12026.05 | 45.78 | |
| Video-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=32026.05 | 45.78 | |
| V-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=32026.05 | 45.78 | |
| Base ModelBackbone=Qwen3-VL-4B-Instruct2026.05 | 45.78 | |
| Video-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=12026.05 | 45.7 | |
| Video-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=22026.05 | 45.7 | |
| Base ModelBackbone=Qwen3-VL-8B-Instruct2026.05 | 45.7 | |
| VisPlayBackbone=Qwen3-VL-8B-Instruct, Iteration=32026.05 | 45.7 | |
| VisPlayBackbone=Qwen3-VL-8B-Instruct, Iteration=42026.05 | 45.7 | |
| V-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=22026.05 | 45.7 | |
| V-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=52026.05 | 45.7 | |
| Base ModelBackbone=Qwen3-VL-8B-Instruct2026.05 | 45.7 | |
| VisPlayBackbone=Qwen3-VL-8B-Instruct2026.05 | 45.7 | |
| V-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=12026.05 | 45.63 | |
| V-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=22026.05 | 45.63 | |
| V-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=32026.05 | 45.63 | |
| VisPlayBackbone=Qwen3-VL-8B-Instruct, Iteration=22026.05 | 45.63 | |
| V-ZeroBackbone=Qwen3-VL-8B-Instruct, Iteration=12026.05 | 45.63 | |
| VisPlayBackbone=Qwen3-VL-4B-Instruct, Iteration=32026.05 | 45.56 | |
| V-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=42026.05 | 45.56 | |
| V-ZeroBackbone=Qwen3-VL-4B-Instruct2026.05 | 45.56 | |
| V-ZeroBackbone=Qwen3-VL-4B-Instruct, Iteration=52026.05 | 45.48 | |
| Video-ZeroBackbone=MiMo-SFT-7B, Checkpoint=v42026.05 | 45.41 | |
| VisPlayBackbone=Qwen3-VL-4B-Instruct, Iteration=22026.05 | 45.33 | |
| VisPlayBackbone=Qwen3-VL-4B-Instruct, Iteration=52026.05 | 45.33 | |
| VisPlayBackbone=Qwen3-VL-4B-Instruct, Iteration=42026.05 | 45.26 | |
| Video-ZeroBackbone=MiMo-SFT-7B, Checkpoint=v32026.05 | 45.26 | |
| VisPlayBackbone=Qwen3-VL-4B-Instruct2026.05 | 45.26 | |
| Video-ZeroBackbone=MiMo-SFT-7B, Checkpoint=v52026.05 | 44.96 | |
| InternVL3.5-4BCheckpoint=Base2026.05 | 38.81 | |
| Video-ZeroBackbone=InternVL3.5-4B, Checkpoint=v12026.05 | 36.89 | |
| Video-ZeroBackbone=InternVL3.5-4B, Checkpoint=v22026.05 | 36.89 | |
| Video-ZeroBackbone=InternVL3.5-4B, Checkpoint=v32026.05 | 36.81 | |
| Video-ZeroBackbone=InternVL3.5-4B, Checkpoint=v42026.05 | 36.67 | |
| Video-ZeroBackbone=InternVL3.5-4B, Checkpoint=v52026.05 | 36.44 |