Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-task Language Understanding on MMLU (Source Lib., Unseen Lib., Delta)
Loading...
75.68
MMLU Source Lib. Accuracy
SIGMA
68.6912
70.5056
72.32
74.1344
Jun 18, 2026
MMLU Source Lib. Accuracy
MMLU Unseen Lib. Accuracy
MMLU Source/Unseen Lib. Delta
Updated 1mo ago
Evaluation Results
Method
Method
Links
MMLU Source Lib. Accuracy
MMLU Unseen Lib. Accuracy
MMLU Source/Unseen Lib. Delta
SIGMA
Base Model=GPT-4o-mini
2026.06
75.68
75.03
0.65
CARD
Base Model=GPT-4o-mini
2026.06
75.11
73.68
1.43
G-Designer
Base Model=GPT-4o-mini
2026.06
73.32
70.05
3.27
GPTSwarm
Base Model=GPT-4o-mini
2026.06
71.37
64.84
6.53
LLM-Debate
Base Model=GPT-4o-mini
2026.06
68.96
63.73
5.23
Feedback
Search any
task
Search any
task