Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multimodal QA on LLaVABench (test)

48LLaVABench Score

PivotMerge

16.38424.59232.841.008Apr 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
48
2026.04
41.5
2026.04
41.3
2026.04
41
2026.04
41
2026.04
40.5
2026.04
37.4
2026.04
37.1
2026.04
36.9
2026.04
33.1
2026.04
33
2026.04
28.2
2026.04
17.6