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

Visual Commonsense Reasoning (Q->A) on VCR sampled 5000 (val)

0.84Accuracy

MerlotReserve

0.22640.38570.5450.7043May 24, 2023
Updated 1mo ago

Evaluation Results

MethodLinks
2023.05
0.84
2023.05
0.708
2023.05
0.638
2023.05
0.507
2023.05
0.406
2023.05
0.283
0.25