Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Scientific Reasoning on GPQA (Acc (%), ECE)

0.3817Accuracy

gemma3-27b-slerp

0.3258520.3403510.354850.369349Oct 20, 2025
Updated 22d ago

Evaluation Results

MethodLinks
2025.10
0.38170.35
2025.10
0.37840.305
2025.10
0.37840.403
2025.10
0.37420.266
2025.10
0.37330.219
2025.10
0.37080.455
2025.10
0.36910.088
2025.10
0.36740.088
2025.10
0.36490.135
2025.10
0.36240.59
2025.10
0.36160.14
2025.10
0.35820.121
2025.10
0.35740.123
2025.10
0.35650.229
2025.10
0.35570.093
2025.10
0.35490.192
2025.10
0.35320.175
2025.10
0.35320.224
2025.10
0.35320.224
2025.10
0.35320.084
2025.10
0.35320.174
2025.10
0.35230.272
2025.10
0.35150.274
2025.10
0.35150.279
2025.10
0.35070.178
2025.10
0.34980.068
2025.10
0.3490.368
2025.10
0.3490.126
2025.10
0.34820.087
2025.10
0.34820.326
2025.10
0.34650.046
2025.10
0.34560.371
2025.10
0.3440.329
2025.10
0.34140.424
2025.10
0.34060.419
2025.10
0.33810.327
2025.10
0.33810.371
2025.10
0.33640.597
2025.10
0.33640.423
2025.10
0.33560.463
2025.10
0.33050.471
2025.10
0.3280.476