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

Malicious Agent on CSQA

0.49ASR@3

DOMINANT

-0.0126320.1178590.248350.378841Dec 21, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
0.490.2711
2025.12
0.460.3022
2025.12
0.440.2978
2025.12
0.42670.5956
2025.12
0.420.5556
2025.12
0.420.5644
2025.12
0.420.5111
2025.12
0.41670.5956
2025.12
0.40670.3289
2025.12
0.39670.5911
2025.12
0.38330.56
2025.12
0.380.5644
2025.12
0.2833-
2025.12
0.26-
2025.12
0.2467-
2025.12
0.2133-
2025.12
0.20330.7333
2025.12
0.12670.7822
2025.12
0.12330.7111
2025.12
0.12330.7467
2025.12
0.08670.96
2025.12
0.05670.96
2025.12
0.04330.9289
2025.12
0.040.96
2025.12
0.03330.9511
2025.12
0.02670.9067
2025.12
0.00670.96
2025.12
0.00670.9556