Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-modal Understanding on MMMU (Specific Accuracy Reduction Ratios)
Loading...
64.3
Accuracy (77.8% reduction ratio)
iLLaVA
57.956
59.603
61.25
62.897
Dec 9, 2024
Accuracy (77.8% reduction ratio)
Accuracy (88.9% reduction ratio)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy (77.8% reduction ratio)
Accuracy (88.9% reduction ratio)
iLLaVA
Style=Token merging
2024.12
64.3
62.8
AIM
Style=Token merging
2024.12
63.8
61.2
VisionZip
Style=Token merging
2024.12
63.6
61.4
FEATHER
Style=Token pruning
2024.12
63.4
61
DivPrune
Style=Token pruning
2024.12
63.1
60.8
FasterVLM
Style=Token pruning
2024.12
62.9
61
AdaFV
Style=Token pruning
2024.12
62.4
59.8
PyramidDrop
Style=Token pruning
2024.12
62.2
57.2
SparseVLM
Style=Token merging
2024.12
62.2
60
LLaVA-PruMerge
Style=Token merging
2024.12
61.8
59.6
FastV
Style=Token pruning
2024.12
58.2
52.4
Feedback
Search any
task
Search any
task
