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

Question Answering on WMDP Cyber QA

44.3Default Accuracy

Base model

23.08428.59234.139.608Jan 31, 2025
Updated 5d ago

Evaluation Results

MethodLinks
2025.01
44.3----
2025.01
44.3----
2025.01
29.127.738.328.540.5
2025.01
29.1-39.528.633.4
2025.01
28.927.94131.739.8
2025.01
28.9-34.528.830.6
2025.01
28.727.138.129.439.4
2025.01
28.7-38.628.434.2
2025.01
28.628.639.830.440.6
2025.01
28.6-39.729.935.5
2025.01
28.227.54031.839.5
2025.01
28.2-35.929.331.7
2025.01
27.928.439.632.641.3
2025.01
27.9-41.432.535.4
2025.01
27.127.442.126.239.9
2025.01
27.1-4326.337.4
2025.01
272742.927.635.7
2025.01
2726.640.332.338.3
2025.01
27-41.22726.1
2025.01
27-37.326.937.7
2025.01
26.726.841.327.337
2025.01
26.7-41.926.627.5
2025.01
2627.442.827.537.3
2025.01
26-41.826.136.8
2025.01
25.827.143.328.139.7
2025.01
25.8-42.527.239.9
2025.01
25.727.542.32639.8
2025.01
25.7-42.42841.4
2025.01
25.627.542.829.840.6
2025.01
25.6-42.527.240.5
2025.01
25.427.542.525.336.3
2025.01
25.4-41.32729.2
2025.01
25.22836.427.436.1
2025.01
25.2-25.32726.2
2025.01
25.127.240.528.935.3
2025.01
25.1-3828.435.3
2025.01
23.926.640.32637.4
2025.01
23.9-36.926.432.4