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

Multilingual Language Understanding on MMMLU (Language Specific Averages)

60.5Accuracy (Korean)

Task Arithmetic

42.09246.87151.6556.429Jan 30, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.01
60.52237.239.9-
2026.01
60.317.437.438.4-
2026.01
60.324.23740.5-
2026.01
60.322.63740-
2026.01
6025.336.640.6-
2026.01
6024.537.740.7-
2026.01
59.924.437.540.6-
2026.01
59.81937.338.7-
2026.01
59.725.137.240.7-
2026.01
59.620.337.339.1-
2026.01
44.528.938.637.3-
2026.01
44.426.137.636-
2026.01
43.625.337.335.4-
2026.01
43.627.537.436.2-
2026.01
43.526.137.535.7-
2026.01
43.424.637.535.2-
2026.01
43.324.837.235.1-
2026.01
43.124.738.235.3-
2026.01
42.927.637.536-
2026.01
42.824.336.734.6-
2026.07
----88.4
2026.07
----86.3
2026.07
----83.4
2026.07
----76.6
2026.07
----67.4
2026.07
----70.7