Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Subject Knowledge Evaluation on MMLU
Loading...
48.7
MMLU Overall Accuracy
Practical (R̄_FPO)
48.388
48.469
48.55
48.631
May 5, 2026
MMLU Overall Accuracy
MMLU STEM Accuracy
MMLU Social Sciences Accuracy
MMLU Humanities Accuracy
MMLU Other Accuracy
Updated 27d ago
Evaluation Results
Method
Method
Links
MMLU Overall Accuracy
MMLU STEM Accuracy
MMLU Social Sciences Accuracy
MMLU Humanities Accuracy
MMLU Other Accuracy
Practical (R̄_FPO)
Evaluation Protocol=50...
2026.05
48.7
41.2
54
53.1
50.6
Relaxed (R̃_FPO)
Evaluation Protocol=50...
2026.05
48.6
41.5
54.3
52.6
49.5
Standard RLHF
Evaluation Protocol=50...
2026.05
48.4
41.3
53.7
52.5
50
Feedback
Search any
task
Search any
task