Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multimodal Reasoning on MMVP (test)
Loading...
0.118
UPR
OpenVLThinker
0.00984
0.03792
0.066
0.09408
Dec 13, 2025
UPR
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
UPR
Accuracy
OpenVLThinker
Mode=Vanilla
2025.12
0.118
0.507
Ocean-R1
Mode=Vanilla
2025.12
0.082
0.413
MM-Eureka
Mode=Vanilla
2025.12
0.069
0.307
MM-Eureka + self-reflection procedure
Mode=Self-reflection
2025.12
0.031
0.373
OpenVLThinker + self-reflection procedure
Mode=Self-reflection
2025.12
0.023
0.477
Ocean-R1 + self-reflection procedure
Mode=Self-reflection
2025.12
0.014
0.417
Feedback
Search any
task
Search any
task