Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Language Understanding on MMLU (Accuracy and Refusal Rate)
Loading...
74.7
MMLU Accuracy
Original
21.868
35.584
49.3
63.016
May 26, 2026
MMLU Accuracy
MMLU Refusal Rate
Updated 7d ago
Evaluation Results
Method
Method
Links
MMLU Accuracy
MMLU Refusal Rate
Original
Model Backbone=Qwen3-14B
2026.05
74.7
-
ICCU (end-to-end)
Model Backbone=Qwen3-14B
2026.05
73.3
2.8
ICCU (filter + generate)
Model Backbone=Qwen3-14B
2026.05
73.1
3.3
O3
Model Backbone=Qwen3-14B
2026.05
61.9
-
Original
Model Backbone=Llama-3...
2026.05
60.6
-
ICCU (filter + generate)
Model Backbone=Llama-3...
2026.05
60.6
3.1
RMU
Model Backbone=Qwen3-14B
2026.05
59.9
-
ICCU (end-to-end)
Model Backbone=Llama-3...
2026.05
58.4
5
O3
Model Backbone=Llama-3...
2026.05
49.8
-
GA
Model Backbone=Llama-3...
2026.05
30.2
-
RMU
Model Backbone=Llama-3...
2026.05
26.4
-
GA
Model Backbone=Qwen3-14B
2026.05
23.9
-
Feedback
Search any
task
Search any
task