Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Model Merging on MMLU, TruthfulQA, BBQ, and CNN/DailyMail
Loading...
69.87
MMLU Score
SA-Merging
67.4676
68.0913
68.715
69.3387
May 30, 2026
MMLU Score
TruthfulQA Score
BBQ Score
CNN/DM Score
Average Score
Updated 1d ago
Evaluation Results
Method
Method
Links
MMLU Score
TruthfulQA Score
BBQ Score
CNN/DM Score
Average Score
SA-Merging
Expert Model=Qwen-14B,...
2026.05
69.87
55.6
81.35
18.48
56.33
TIES Merging
Expert Model=Qwen-14B,...
2026.05
69.38
52.03
81.06
15.91
54.62
WUDI-Merging
Expert Model=Qwen-14B,...
2026.05
69.17
55.71
80.56
17.33
55.69
Individual
Expert Model=Qwen-14B,...
2026.05
68.35
53.34
93.53
19.46
58.67
Task Arithmetic
Expert Model=Qwen-14B,...
2026.05
67.56
52.33
78.38
20.54
54.7
Feedback
Search any
task
Search any
task