Visual Question Answering on TextVQA (Accuracy, Avg. Ratio)
58.2AccuracyVanilla
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| VanillaSource=CVPR’24, Average Token=576, Reroute=false2026.06 | 58.2 | 100 | |
| Nüwa+RerouteSource=Ours, Average Token=192, Reroute=true2026.06 | 57.8 | 100.6 | |
| NüwaSource=ICLR’26, Average Token=192, Reroute=false2026.06 | 57.4 | 99.2 | |
| VisionZipSource=CVPR’25, Average Token=192, Reroute=false2026.06 | 57.3 | 97.9 | |
| Nüwa+RerouteSource=Ours, Average Token=128, Reroute=true2026.06 | 57.3 | 99.8 | |
| NüwaSource=ICLR’26, Average Token=128, Reroute=false2026.06 | 57 | 98.4 | |
| VisionZipSource=CVPR’25, Average Token=128, Reroute=false2026.06 | 56.8 | 96.9 | |
| PDropSource=CVPR’25, Average Token=192, Reroute=false2026.06 | 56.1 | 95.7 | |
| SparseVLMSource=ICML’25, Average Token=192, Reroute=false2026.06 | 56.1 | 95.4 | |
| VisionZipSource=CVPR’25, Average Token=64, Reroute=false2026.06 | 55.5 | 93.2 | |
| PDropSource=CVPR’25, Average Token=128, Reroute=false2026.06 | 55.1 | 93.8 | |
| Nüwa+RerouteSource=Ours, Average Token=64, Reroute=true2026.06 | 55.1 | 96.7 | |
| SparseVLMSource=ICML’25, Average Token=128, Reroute=false2026.06 | 54.9 | 92.9 | |
| NüwaSource=ICLR’26, Average Token=64, Reroute=false2026.06 | 54.9 | 95.8 | |
| PruMergeSource=ICCV’25, Average Token=128, Reroute=false2026.06 | 54.3 | 94.7 | |
| SparseVLMSource=ICML’25, Average Token=64, Reroute=false2026.06 | 53.4 | 91.5 | |
| FastVSource=ECCV’24, Average Token=192, Reroute=false2026.06 | 52.5 | 90.1 | |
| PruMergeSource=ICCV’25, Average Token=64, Reroute=false2026.06 | 52 | 92.1 | |
| FastVSource=ECCV’24, Average Token=128, Reroute=false2026.06 | 50.6 | 85.2 | |
| FastVSource=ECCV’24, Average Token=64, Reroute=false2026.06 | 47.8 | 77.9 | |
| PDropSource=CVPR’25, Average Token=64, Reroute=false2026.06 | 45.9 | 70.3 |