Multimodal Question Answering on MM-Lifelong week (test)
95.6AccuracyHuman
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| HumanFrames=Full2026.03 | 95.6 | 42.4 | |
| ReMAFrames=Full2026.03 | 18.82 | 16.37 | |
| Qwen3-VL-235B-A22BFrames=15362026.03 | 15.63 | 0.8 | |
| GPT-5Frames=502026.03 | 15 | 0.92 | |
| Video-XL-2-8BFrames=10242026.03 | 12 | 0.51 | |
| VideoMind-7BFrames=Full2026.03 | 11.75 | 2.51 | |
| Qwen3-VL-30B-A3BFrames=15362026.03 | 11.07 | 0.77 | |
| Nemotron-v2-12BFrames=5122026.03 | 11 | 0.5 | |
| Video-XL-2-8BFrames=20482026.03 | 10.25 | 0.1 | |
| LongVT-7BFrames=Full2026.03 | 9.75 | 0.66 | |
| Eagle-2.5-8BFrames=5122026.03 | 9.5 | 1.69 | |
| DeepVideoDiscoveryFrames=Full2026.03 | 9.02 | 8.12 | |
| Nemotron-v2-12BFrames=1282026.03 | 8.5 | 0.5 | |
| Eagle-2.5-8BFrames=322026.03 | 7 | 1.16 |