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

Multimodal Evaluation on MMMU

57.6Score

Vanilla

Updated 19d ago

Evaluation Results

Method	Links
Vanilla 2026.05		57.6
EVA 2026.05		57.6
Vanilla 2026.05		57.6
EVA 2026.05		57.6
VLGuard 2026.05		57.1
VLGuard 2026.05		57.1
EVA(text) 2026.05		57
EVA(text) 2026.05		57
JailGuard 2026.05		55.8
JailGuard 2026.05		55.8
VLGuard 2026.05		50.2
VLGuard 2026.05		50.2
Vanilla 2026.05		49.9
Vanilla 2026.05		49.9
EVA(text) 2026.05		49.6
EVA(text) 2026.05		49.6
EVA 2026.05		48.2
EVA 2026.05		48.2
JailGuard 2026.05		47.5
JailGuard 2026.05		47.5
Adashield-A 2026.05		38.4
Adashield-A 2026.05		38.4
Adashield-A 2026.05		35.4
Adashield-A 2026.05		35.4
EVA 2026.05		33.4
EVA 2026.05		33.4
EVA(text) 2026.05		33.3
EVA(text) 2026.05		33.3
VLGuard 2026.05		32.1
VLGuard 2026.05		32.1
Vanilla 2026.05		31.3
Vanilla 2026.05		31.3
JailGuard 2026.05		29.8
JailGuard 2026.05		29.8
Adashield-A 2026.05		19.7
Adashield-A 2026.05		19.7