Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Scientific Reasoning on GPQA (Accuracy, CLC)
Loading...
64.96
Accuracy
SOLAR
25.8352
35.9926
46.15
56.3074
Jun 25, 2026
Accuracy
CLC
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
CLC
SOLAR
Backbone=Qwen3-8B, Fin...
2026.06
64.96
57.27
SFT
Backbone=Qwen3-8B, Fin...
2026.06
63.64
55.73
Soft Thinking
Backbone=Qwen3-8B, Fin...
2026.06
53.06
43.42
Qwen3-8B
Backbone=Qwen3-8B, Fin...
2026.06
52.68
42.69
SOLAR
Backbone=Qwen3-4B, Fin...
2026.06
45.24
36.26
SFT
Backbone=Qwen3-4B, Fin...
2026.06
43.75
33.15
AlignX
Backbone=Qwen3-4B, Fin...
2026.06
43.03
32.97
MidAlign
Backbone=Qwen3-4B, Fin...
2026.06
43
32.47
MAPO
Backbone=Qwen3-4B, Fin...
2026.06
40.32
30.15
Qwen3-4B
Backbone=Qwen3-4B, Fin...
2026.06
27.52
14.56
Soft Thinking
Backbone=Qwen3-4B, Fin...
2026.06
27.34
14.31
Feedback
Search any
task
Search any
task