Video Question Answering on PerceptionTest (val)
66.9Validation ScoreFull Tokens
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Full TokensBackbone=LLaVA-OV-72B2024.11 | 66.9 | — | |
| DyCokeBackbone=LLaVA-OV-72B, K=0.5, L=3, P=0.72024.11 | 60.2 | — | |
| DyCokeBackbone=LLaVA-OV-72B, K=0.7, L=3, P=0.72024.11 | 59.6 | — | |
| FastVBackbone=LLaVA-OV-72B2024.11 | 58.5 | — | |
| DyCokeBackbone=LLaVA-OV-7B, K=0.5, L=3, P=0.72024.11 | 57.6 | — | |
| DyCokeBackbone=LLaVA-OV-7B, K=0.7, L=3, P=0.72024.11 | 57.6 | — | |
| DyCokeBackbone=LLaVA-OV-7B, K=0.3, L=3, P=0.72024.11 | 57.2 | — | |
| Full TokensBackbone=LLaVA-OV-7B2024.11 | 57.1 | — | |
| FastVBackbone=LLaVA-OV-7B2024.11 | 56.7 | — | |
| PruMergeBackbone=LLaVA-OV-72B2024.11 | 55.6 | — | |
| PruMergeBackbone=LLaVA-OV-7B2024.11 | 54.3 | — | |
| DyCokeBackbone=LLaVA-OV-0.5B, K=0.5, L=3, P=0.72024.11 | 49.5 | — | |
| DyCokeBackbone=LLaVA-OV-0.5B, K=0.7, L=3, P=0.72024.11 | 49.5 | — | |
| Full TokensBackbone=LLaVA-OV-0.5B2024.11 | 49.1 | — | |
| FastVBackbone=LLaVA-OV-0.5B2024.11 | 49.1 | — | |
| DyCokeBackbone=LLaVA-OV-0.5B, K=0.3, L=3, P=0.72024.11 | 48.9 | — | |
| PruMergeBackbone=LLaVA-OV-0.5B2024.11 | 47.5 | — | |
| Apollo-7BParameters=7B2026.02 | — | 67.3 | |
| CoPE-VideoLM-7BParameters=7B2026.02 | — | 70.5 | |
| IXC-2.5-7BParameters=7B2026.02 | — | 34.4 | |
| LLaVA-OV-7BParameters=7B2026.02 | — | 57.1 | |
| LLaVA-Video-7BParameters=7B2026.02 | — | 67.9 | |
| Oryx-7BParameters=7B2026.02 | — | 68.6 | |
| Video-LaVIT2026.02 | — | 47.9 | |
| VILA-40BParameters=40B2026.02 | — | 54 |