List your company

Multitask Language Understanding on M_MMLU_c (accuracy)

30.07Accuracy (MMLU-c)ML (Q3)
28.239628.714829.1929.6652Apr 22, 2026

Evaluation results

MethodLinks
ML (Q3)Filtering Strategy=ML (Q3)2026.0430.07
Q3 negativesFiltering Strategy=Q3 negatives2026.0429.92
ML (15%)Filtering Strategy=ML (15%)2026.0429.91
HQ Romance (No Fra)Filtering Strategy=HQ Romance (No Fra)2026.0429.5
HQ allFiltering Strategy=HQ all2026.0429.49
HQ seed (LLM Training)Filtering Strategy=HQ seed (LLM Training)2026.0429.46
HQFiltering Strategy=HQ2026.0429.45
HQ seedFiltering Strategy=HQ seed2026.0429.44
NordicFiltering Strategy=Nordic2026.0429.36
MKC-eFiltering Strategy=MKC-e2026.0429.29
MLFiltering Strategy=ML2026.0429.27
No FilteringFiltering Strategy=No Filtering2026.0428.31