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

Procedural Multimodal Machine Comprehension on CraftQA (test)

55.33Cloze

Human

26.709234.139641.5749.0004Apr 6, 2022
Updated 5mo ago

Evaluation Results

MethodLinks
2022.04
55.3361.854.657.24
2022.04
48.0151.3649.2549.54
2022.04
47.4950.0348.7948.77
2022.04
46.6749.0648.3548.02
2022.04
45.9147.7147.2446.95
2022.04
44.8746.2946.7745.98
2022.04
40.5647.8446.1844.86
2022.04
39.4346.5243.6243.2
2022.04
35.1433.6742.3137.04
2022.04
27.8142.6135.2335.22