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

General Knowledge on MMLU (MMLU, UtilityNorm, Score, Rank)

64MMLU Accuracy

Original

Updated 19d ago

Evaluation Results

Method	Links
Original 2025.09		64	-	-	-
DOOR 2025.09		55.5	98.5	28.9	5
Task Vector 2025.09		53.3	100	0	8
GAGDR 2025.09		48	92.1	52.8	4
NPOGDR 2025.09		46.3	80.6	55.6	3
PRISM 2025.09		38.5	73.2	76.1	1
SAM+NPO 2025.09		23.8	53	72.1	2
NPO 2025.09		23	0.3	5.5	6
GA 2025.09		22.9	0	0	7