ML//benchmark//MMLU

MMLU (Massive Multitask Language Understanding) is a multiple-choice benchmark covering 57 subjects, from elementary math to professional law. GPT-4 scores about 86%, approaching human expert level.

MMLU (Massive Multitask Language Understanding) is a multiple-choice benchmark covering 57 subjects, from elementary math to professional law. GPT-4 scores about 86%, approaching human expert level.

It carries heavy contamination concerns, because some of its questions appear verbatim in training data.