Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-image Dialogue Understanding on MMDU
Loading...
26.37
Accuracy
LLaVA-v1.5
9.7612
14.0731
18.385
22.6969
Oct 22, 2025
Accuracy
Updated 26d ago
Evaluation Results
Method
Method
Links
Accuracy
LLaVA-v1.5
Backbone Parameters=7B
2025.10
26.37
KORE
rank=256
2025.10
23.18
Replay
2025.10
19.31
MoELoRA
2025.10
17.85
O-LoRA
2025.10
17.28
KORE
rank=235
2025.10
16.58
LoRA
2025.10
13.7
LwF
2025.10
13.68
EWC
2025.10
13.36
Full-FT
Learning Strategy=Fine...
2025.10
13.03
SEFE
2025.10
10.55
CIA
2025.10
10.4
Feedback
Search any
task
Search any
task