Video Question Answering on LLaVA-Video
63.2VYRQwen3-VL-30B
Evaluation Results
| Method | Links | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Qwen3-VL-30BParameters=30B2026.04 | 63.2 | 71.4 | 72.5 | 3.807 | 27.6 | 6.6 | 80.4 | 69.9 | 78.5 | 0.523 | 2.509 | 2.422 | 2.145 | 50.4 | |
| Baseline (Qwen3-VL-8B-Instruct)Backbone=Qwen3-VL-8B-Instruct2026.04 | 59.5 | 71.1 | 73 | 3.853 | 27.8 | 16.3 | 79.6 | 68.2 | 81.9 | 0.18 | 2.223 | 3.31 | 2.361 | 29.3 | |
| Baseline+SFTBackbone=Qwen3-VL-8B-Instruct, Supervised Fine-Tuning=true2026.04 | 57 | 67.6 | 69.4 | 3.851 | 31 | 13.9 | 77.7 | 64.8 | 77.2 | 0.178 | 2.435 | 3.028 | 2.399 | 51.4 | |
| Baseline+CoTBackbone=Qwen3-VL-8B-Instruct, Chain-of-Thought=true2026.04 | 55 | 68.6 | 70.2 | 3.843 | 29.5 | 11.4 | 77.6 | 64.8 | 81.1 | 0.498 | 2.585 | 3.541 | 1.462 | 50.7 | |
| Internvl3.5-VL-241BParameters=241B2026.04 | 52.4 | 62.5 | 63.8 | 3.791 | 49.7 | 4.9 | 78.3 | 63.5 | 80.3 | 0.879 | 2.816 | 2.382 | 1.894 | 52.8 | |
| Qwen3-VL-235BParameters=235B2026.04 | 49.6 | 60.1 | 61.3 | 3.785 | 50.2 | 3.4 | 76.7 | 60.3 | 69.4 | 0.796 | 2.743 | 2.223 | 1.547 | 55.3 | |
| Baseline+CoT+SFTBackbone=Qwen3-VL-8B-Instruct, Chain-of-Thought=true, Supervised Fine-Tuning=true2026.04 | 30.9 | 40.7 | 42.2 | 3.892 | 55.9 | 15.8 | 60.4 | 39.4 | 46.6 | 0.287 | 2.671 | 1.869 | 1.482 | 69.7 | |
| VTHM-MoE2026.04 | 4.6 | 8.7 | 8.6 | 3.722 | 77.1 | 19 | 35.8 | 0.3 | 2.2 | 1.664 | 0.765 | 1.781 | 0.983 | 77.7 | |
| Gemini-3.1-Pro2026.04 | 1.1 | 7.4 | 7.1 | 3.629 | 72.8 | 32.2 | 33.5 | 1.5 | 0.3 | 1.027 | 0.287 | 2.161 | 0.155 | 73.7 |