Audio-Visual Reasoning on WorldSense
54.3ScoreFull tokens
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Full tokensR (Rv-Ra)=100 (100-100), TFLOPs=47.22026.05 | 54.3 | 55.5 | 100 | |
| SEATSR (Rv-Ra)=35 (32-70), TFLOPs=13.32026.05 | 54.1 | 55.4 | 99.8 | |
| OmniZipR (Rv-Ra)=35 (32-70), TFLOPs=13.42026.05 | 53.8 | 54.6 | 98.4 | |
| SEATSR (Rv-Ra)=25 (22-60), TFLOPs=9.02026.05 | 53.6 | 54.7 | 98.6 | |
| DivPruneR (Rv-Ra)=35 (29-100), TFLOPs=13.42026.05 | 53.4 | 54 | 97.3 | |
| VisionZip-omR (Rv-Ra)=35 (32-70), TFLOPs=13.42026.05 | 53.3 | 54.5 | 98.2 | |
| DivPrune-omR (Rv-Ra)=35 (32-70), TFLOPs=13.42026.05 | 53.2 | 54.2 | 97.7 | |
| VisionZipR (Rv-Ra)=35 (29-100), TFLOPs=13.42026.05 | 53 | 54.4 | 98 | |
| DyCokeR (Rv-Ra)=35 (29-100), TFLOPs=13.42026.05 | 52.9 | 53.7 | 96.7 | |
| RandomR (Rv-Ra)=35 (32-70), TFLOPs=13.42026.05 | 52.6 | 53.6 | 96.6 | |
| DivPrune-omR (Rv-Ra)=25 (22-60), TFLOPs=9.42026.05 | 52.2 | 53.7 | 96.7 | |
| DivPruneR (Rv-Ra)=25 (18-100), TFLOPs=9.42026.05 | 52.1 | 53.6 | 96.6 | |
| VisionZipR (Rv-Ra)=25 (18-100), TFLOPs=9.42026.05 | 51.9 | 52.9 | 95.3 | |
| SEATSR (Rv-Ra)=15 (12-50), TFLOPs=5.52026.05 | 51.9 | 53.8 | 96.9 | |
| VisionZip-omR (Rv-Ra)=25 (22-60), TFLOPs=9.42026.05 | 51.8 | 53.8 | 96.9 | |
| FastVR (Rv-Ra)=35 (29-100), TFLOPs=13.72026.05 | 51.5 | 53.3 | 96 | |
| SEATSR (Rv-Ra)=10 (7-45), TFLOPs=3.92026.05 | 51.5 | 53 | 95.5 | |
| FastVIDR (Rv-Ra)=25 (18-100), TFLOPs=9.12026.05 | 51.4 | 52.9 | 95.3 | |
| RandomR (Rv-Ra)=25 (22-60), TFLOPs=9.42026.05 | 51.3 | 52.9 | 95.3 | |
| DivPruneR (Rv-Ra)=15 (8-100), TFLOPs=5.82026.05 | 51 | 51.8 | 93.3 | |
| OmniZipR (Rv-Ra)=25 (25-25), TFLOPs=8.52026.05 | 50.7 | 52.5 | 94.5 | |
| DivPrune-omR (Rv-Ra)=15 (12-50), TFLOPs=5.82026.05 | 50.7 | 51.8 | 93.3 | |
| VisionZip-omR (Rv-Ra)=15 (12-50), TFLOPs=5.82026.05 | 50.2 | 51.5 | 92.8 | |
| RandomR (Rv-Ra)=15 (12-50), TFLOPs=5.82026.05 | 49.8 | 50.6 | 91.2 | |
| VisionZipR (Rv-Ra)=15 (8-100), TFLOPs=5.82026.05 | 49.8 | 51.2 | 92.3 | |
| FastV-omR (Rv-Ra)=35 (32-70), TFLOPs=13.72026.05 | 49.7 | 51.4 | 92.6 | |
| FastVR (Rv-Ra)=25 (18-100), TFLOPs=9.82026.05 | 49.5 | 51.6 | 93 | |
| DivPrune-omR (Rv-Ra)=10 (7-45), TFLOPs=4.02026.05 | 49.3 | 50.9 | 91.7 | |
| VisionZip-omR (Rv-Ra)=10 (7-45), TFLOPs=4.02026.05 | 48.9 | 49.8 | 89.7 | |
| RandomR (Rv-Ra)=10 (7-45), TFLOPs=4.02026.05 | 48.3 | 49.1 | 88.5 | |
| FastV-omR (Rv-Ra)=25 (22-60), TFLOPs=9.82026.05 | 47.2 | 49.5 | 89.2 | |
| FastV-omR (Rv-Ra)=15 (12-50), TFLOPs=6.32026.05 | 44.6 | 46.7 | 84.2 |