List your company

Discriminative Hallucination Evaluation on AMBER (Acc, P, R, F1)

84.3AccuracyMESA
66.30870.97975.6580.321Apr 4, 2026Apr 6, 2026Apr 9, 2026Apr 12, 2026Apr 14, 2026Apr 17, 2026Apr 20, 2026

Evaluation results

MethodLinks
MESA2026.0484.390.385.487.8
Nullu2026.0481.587.484.285.8
PSRDBase Model=LLaVA-1.5-7B2026.0481.2——85
ICT2026.0479.584.784.484.5
VTI2026.0479.182.187.684.8
OctopusBase Model=LLaVA-1.5-7B2026.0476.7——82.7
ICD2026.0474.173.296.283.1
VCD2026.0473.983.875.279.3
Shikra-7BVariant=Original2026.0473.585.772.878.7
Shikra-7BVariant=Ours2026.0473.385.174.279.3
Vanilla2026.0471.481.973.177.2
LLaVA-1.5-7BVariant=Ours2026.0471.189.563.974.6
AVISCBase Model=LLaVA-1.5-7B2026.0470.7——75.5
LLaVA-1.5-7BVariant=Original2026.04709061.773.2
VAF2026.0469.894.857.671.7
VCDBase Model=LLaVA-1.5-7B2026.0467.3——71.1
M3IDBase Model=LLaVA-1.5-7B2026.0467.3——70.9
LLaVA-1.5-7B (Base)2026.0467——71.1