Visual Question Answering on Open Images cross-task (test)
44.7AccuracyEVPG
Evaluation Results
| Method | Links | |
|---|---|---|
| EVPGMethod Category=VisProg, Evaluation Protocol=Supervised, Backbone=InternVL1.52025.12 | 44.7 | |
| InternVL1.5 (SFT)Method Category=End-to-end, Evaluation Protocol=Supervised2025.12 | 42.7 | |
| InternVL1.5 (SDVP)Method Category=VisProg, Evaluation Protocol=Supervised2025.12 | 42.5 | |
| InternVL1.5Method Category=End-to-end, Evaluation Protocol=Zero-shot2025.12 | 41 | |
| InternVL1.5Method Category=VisProg, Evaluation Protocol=Zero-shot2025.12 | 40.6 |