Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Multistep Soft Reasoning on MUSR (Accuracy %)
Loading...
43.1
Accuracy (%)
Qwen 3 8B
37.484
38.942
40.4
41.858
Sep 26, 2025
Accuracy (%)
Updated 4d ago
Evaluation Results
Method
Method
Links
Accuracy (%)
Qwen 3 8B
Compression Ratio=Full
2025.09
43.1
CoSpaDi
Base Model=Qwen 3 8B,...
2025.09
42.1
SVDLLM
Base Model=Qwen 3 14B,...
2025.09
41.5
SVDLLM
Base Model=Qwen 3 8B,...
2025.09
41.4
Qwen 3 14B
Compression Ratio=Full
2025.09
40.7
CoSpaDi
Base Model=Qwen 3 14B,...
2025.09
40.7
SVDLLM
Base Model=Qwen 3 8B,...
2025.09
39.8
SVDLLM
Base Model=Qwen 3 14B,...
2025.09
39.7
CoSpaDi
Base Model=Qwen 3 8B,...
2025.09
38.4
CoSpaDi
Base Model=Qwen 3 8B,...
2025.09
38.1
CoSpaDi
Base Model=Qwen 3 14B,...
2025.09
38
SVDLLM
Base Model=Qwen 3 8B,...
2025.09
37.7
Feedback
Search any
task
Search any
task