Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Multimodal In-Context Learning on VL-ICL benchmark

77.6Average Accuracy

MM1.5-30B

10.10427.62745.1562.673Sep 30, 2024
Updated 3d ago

Evaluation Results

MethodLinks
2024.09
77.646.566.5100658044.5
2024.09
65.8428156927450
2024.09
59.632589263.36547.5
2024.09
56.333.5598848.366.742.5
2024.09
56.3------
2024.09
563356.58956.756.744
2024.09
5625.552.898.568.36031
2024.09
56------
2024.09
52.83369.597.5404532
2024.09
52.1256398.551.738.336
2024.09
513952846036.734
2024.09
51------
2024.09
3727.5507918.313.334
2024.09
35.631.55052.521.728.329.5
2024.09
34.32549.37316.78.333.5
2024.09
31.926.856.45818.98.922.3
2024.09
30.92050.14939.411.116
2024.09
2827.75053.87.86.122.8
2024.09
26.914.85028.221.71036.5
2024.09
20.918.85051.22.82.80
2024.09
19.51750126.78.314
2024.09
19.5------
2024.09
18.5------
2024.09
17.98.250.428.512.27.20.8
2024.09
12.717.85003.350