Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multimodal Reasoning on SEED
Loading...
32.52
ECE
MergeMix
32.3244
33.6447
34.965
36.2853
Oct 27, 2025
ECE
Updated 1mo ago
Evaluation Results
Method
Method
Links
ECE
MergeMix
Vision Token Count=Ful...
2025.10
32.52
SFT Vision
Vision Token Count=Full
2025.10
32.67
CutMix
Vision Token Count=Ful...
2025.10
32.77
MixUp
Vision Token Count=Ful...
2025.10
33.26
MergeMix
Vision Token Count=50%...
2025.10
33.37
MixUp
Vision Token Count=50%...
2025.10
33.6
Baseline
2025.10
33.79
CutMix
Vision Token Count=50%...
2025.10
33.83
SFT Vision
Vision Token Count=50%
2025.10
34.41
ResizeMix
Vision Token Count=Ful...
2025.10
36.08
ResizeMix
Vision Token Count=50%...
2025.10
37.41
Feedback
Search any
task
Search any
task
