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

Multilingual Language Understanding on MMLU-ProX (test)

43.6Accuracy

COMPASS

17.49624.27331.0537.827Apr 22, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
43.6
2026.04
42.5
2026.04
39.4
2026.04
39
2026.04
38.7
2026.04
38.4
2026.04
37.5
2026.04
33.7
2026.04
28.7
2026.04
27.7
2026.04
26.9
2026.04
26.1
2026.04
24.5
2026.04
24.4
2026.04
24.2
2026.04
24
2026.04
23.8
2026.04
23.7
2026.04
23.2
2026.04
22.9
2026.04
22.8
2026.04
21.7
2026.04
19.6
2026.04
18.5